When a person has been diagnosed with a terminal illness and has less than 6 months to live, with curative treatments no longer being viable, end-of-life care or hospice care is initiated (Huffman and Harmer, 2022). It is estimated that approximately 7 million people per year receive this type of care (Connor et al, 2021). The aim of this care is to reduce distress and improve the quality of life (QoL) for patients and their families by providing comprehensive support with physical, psychosocial, and spiritual needs of the individual (Rome et al, 2011). Survey data suggest that, when given the choice, most individuals prefer to receive end-of-life care at home (Arnold et al, 2015; Higginson et al, 2017; Ali et al, 2019). However, even when end-of-life care is provided, there are still uncertainties regarding the effect on the risk of likelihood of dying in hospital, and what effect providing this service has on patients' symptoms and QoL (Shepperd et al, 2021). Subsequently a Cochrane review was conducted/updated to assess the effects of receiving end-of-life care at home on these outcomes (Shepperd et al, 2021).
Aim
This commentary aims to critically appraise the methods used within the review by Shepperd et al (2021) and to expand upon its findings in the context of clinical practice.
Methods
A comprehensive multi-database search was undertaken from date of inception until March 2020. No restrictions on publication type, or language were applied to the search. In addition, forward (using ISI Web of Science) and backward citation searches were undertaken for all included studies. Additional hand-searching was also undertaken. Only randomised controlled trials (RCTs) or cluster random controlled trials, which included adults who received end-of-life care at home compared with inpatient hospital or hospice care, were included. End-of-life care at home was defined as a service offered by healthcare professionals, which provides active treatment for extended periods in the patient's home, thus avoiding the need for hospital or hospice inpatient end-of-life care. Title and abstract screening were undertaken by a single reviewer. Full paper screening, data extraction and risk of bias (Cochrane Risk of Bias Tool) was carried out by two reviewers independently. Additional evaluation of confidence in the estimation of effect was undertaken using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) criteria (Guyatt et al, 2008). When feasible, a fixed-effect Mantel-Haenszel meta-analysis was undertaken for dichotomous data to calculate a risk ratio (RR) and corresponding 95% confidence interval (CI 95%). To assess the variance in observed effects rather than sampling error, the I2 statistic was assessed. To test the null hypothesis that all studies in the analysis share a common effect size, the Q statistic was utilised.
Results
For this current update, 2984 citations were identified after duplicate removal. After screening, no additional papers were identified, resulting in four RCTs being included in this Cochrane systematic review. Of these four RCTs two were conducted in the US, one in England and one in Norway. Three out of the four RCTs used a multidisciplinary team approach, with the remaining RCTs using a single disciplinary nursing care approach.
When home-based end-of-life care was provided, it increased the risk of the place of death being at home (RR 1.31; 95% CI, 1.12 to 1.52; GRADE: High) compared to combined services (this could include routine (not specialised) home care, acute inpatient care, primary care services and hospice care). Low quality evidence from two RCTs indicated that there was a cost reduction in healthcare services for end-of-life care at home of between 18 to 30%.
For the outcome of unplanned admission to hospital (GRADE: Low) there was a wide variation of individual study estimates of relative risk ranging from 2.61 (95% CI, 1.50 to 4.55) to 0.62 (95% CI, 0.48 to 0.79). Low-quality evidence demonstrated similar inconsistencies regarding the effect of home-based end-of-life care on combined services on satisfaction. One RCT reported a small increase in satisfaction and one reported no evidence of difference. Very low-quality evidence suggests that there could be a small reduction in pain control assessed by caregivers (0.48 points, 95% CI –0.93 to –0.03, four-point scale). There is very low-quality evidence from one RCT suggesting that caregivers of participants receiving home-based end-of-life care exhibited a decrease in psychological well-being and a little or no difference in the questionnaire assessing bereavement. District nurses reported within one RCT (very low quality) that there was less of a need for carer support (1.36 vs 1.81, P ≤ 0.01) and additional help with night nursing (1.43 vs 2.03, P < 0.0001) in the home-based end-of-life care group compared to control (three-point scale negative scores represent less of a problem).
Commentary
Using the A MeaSurement Tool to Assess systematic Reviews(AMSTAR) 2 critical appraisal tool for systematic reviews, 14 out of 16 criteria were judged to be satisfactory for this review (Shea et al, 2017). The only two criteria that were not achieved was that the authors failed to explain why only RCTs were used and did not conduct duplicate screening for titles and abstracts. Due to the limited evidence within this area, taking a more expansive approach (including non-randomised studies) may provide a more comprehensive overview of the current evidence base on the effects of home-based end-of-life care. However, when including non-random controlled trials within a systematic review, it can lead to additional bias regarding sample allocation and affect the certainty in any type of recommendation being made (Deeks et al, 2003). Regarding title and abstract screening, there is a possibility that studies were missed during the screening process, as title and abstract screening were undertaken by a single reviewer (Waffenschmidt et al, 2019). Overall, the systematic review provides a comprehensive synthesis of the included studies.
The findings from this review indicate with high certainty that home-based end-of-life care may increase the likelihood that the place of death is at home. Based upon these estimates, it would result in an increase of 163 deaths per 1000 deaths occurring at the individual's place of home when receiving home-based end-of-life care, compared to combined services. However, it is important to note that dying at home is not the only outcome to indicate good end-of-life care (Age UK, 2019). Unfortunately, the other outcomes assessed in the review produced notable uncertainties, which makes it difficult to suggest any specific recommendations for practice, based upon this evidence alone. There was very low-quality evidence, suggesting that home-based end-of-life care may provide small improvement in pain management assessed by care givers. When care givers are required to deliver pain relief, qualitative evidence suggests that care givers need specialist training on pain management (Chi and Demiris, 2017). This training should focus on individual knowledge related to pain management, assessment, interpretation and use of pain medications (Chi and Demiris, 2017). The training may be provided through face-to-face education supported with written or multimedia resources (Latter et al, 2016).
Regarding the support of carers, this Cochrane systematic review found that there was very low evidence that home-based end-of-life care resulted in the perception that such intervention required less carer support than the control group. However, this was based upon a three-point scale and despite this being a significant difference, the actual means difference was minimal, suggesting that both groups could require additional support (Grande et al, 2000). As highlighted in the National Institute for Health and Care Excellence (NICE) guideline on end-of-life care for adults, the emotional and practical needs of carers should be evaluated and reassessed when required (NICE, 2019). These needs are often multifaceted, in areas such as emotional support, knowledge related to the disease, taking on responsibilities as a career, caring for themselves, and general practical assistance (Marco et al, 2022). These needs can be met by providing access to professional services, communication with health professionals and educational opportunities (Marco et al, 2022). Finally, the review suggested that receiving home-based end-of-life care may result in reduced service costs of between 18-30% compared to combined services. However, this is based upon low certainty evidence; therefore, these estimates should be viewed with caution.
End-of-life care at home vs hospital is a difficult decision for patients and their relatives, and can present a significant source of confusion and anxiety (Gomes and Higginson, 2004). Future research should focus on aspects that would help patients and their relatives when making this decision. Further exploration should be undertaken focusing on the effects of home-based end-of-life care, on frequency and type of unplanned admissions to hospitals, which can disrupt continuity of care and patients' end-of-life goals. Another avenue of focus for future research should be on effects of home-based end-of-life care on patient-specific outcomes. There is currently no overall core outcome set for end-of-life care (Zambrano et al, 2020). However, one is currently being developed and when published, the recommendation should be followed regarding which outcomes are important for patients (Zambrano et al, 2020). Future research should investigate the effect of the hospital-at-home program on various subgroups of patients with terminal illnesses or distinct demographics, with the aim of providing personalized end-of-life care to particular patient groups.
Key points
- High quality evidence suggests that end-of-life care at home may increase the risk of an individual dying in their place of home compared to combined services
- Low quality evidence suggests that end-of-life care at home may have reduced costs compared to combined services
- There is still much uncertainty regarding the effect of end-of-life care at home on unplanned admissions to hospital, caregiver outcomes and staff views on the provision of services
- Future research should focus on home-based end-of-life care programs in different patient populations and provide more information on important aspects, such as unplanned admissions to hospital and health outcomes, to facilitate easier decision making for patients and relatives.
CPD reflective questions
- What can home-based end-of-life care provide for patients?
- What are the benefits of using the Cochrane Risk of Bias tool?
- What still needs to be identified before home-based end-of-life care can be recommended?