Can We Make More Accurate Prognoses During Last Days of Life?

Background: Life expectancy prediction is important for end-of-life planning. Established methods (Palliative Performance Scale [PPS], Palliative Prognostic Index [PPI]) have been validated for intermediate- to long-term prognoses, but last-weeks-of-life prognosis has not been well studied. Patients admitted to a palliative care facility often have a life expectancy of less than three weeks. Reliable last-weeks-of-life prognostic tools are needed. Objective: To improve short-term survival prediction in terminally ill patients. Method: This prospective study included all patients admitted to a palliative care facility in Montreal, Canada, over one year. PPS and PPI were assessed until patients' death. Seven prognostic clinical signs of impending death (Short-Term Prognosis Signs [SPS]) were documented daily. Results: The analyses included 273 patients (76% cancer). The median survival time for a PPS ≤20% was 2.5 days, while for a PPS ≥50% it was 44.5 days, for a PPI >8 the median survival was 3.5 days and for a PPI ≤4 it was 38.5 days. Receiver operating characteristic curves showed a high accuracy in predicting survival. Median survival after the first occurrence of any SPS was below one week. Conclusions: This study demonstrated that the PPS and PPI perform well between one week and three months extending their usefulness to shorter term survival prediction. SPS items provided survival information during the last week of life. Using SPS along with PPS and PPI during the last weeks of life could enable a more precise short-term survival prediction across various end-of-life diagnoses. The translation of this research into clinical practice could lead to a better adapted treatment, the identification of a most appropriate care setting for patients, and improved communication of prognosis with patients and families.

A Prospective Study with Patients and Families on the Usefulness of Accurate Prognosis for Palliative Care Patients.

Prediction of life expectancy in terminally ill patients is an important end-of-life care issue for patients, families and mental health workers during the last days of life. This study was conducted to examine the importance/usefulness for patients/families to have an accurate prognosis and its impact on planning their activities prior to death. All patients admitted during a period of one year were included. Patients' and families' viewpoints on the usefulness of an accurate prognosis was documented at admission. There were 285 patients in the cohort. The median time to death was 8 days. Most families (83%) rated the importance of an accurate prognosis as moderately (13%) to very much useful (70%). A total of 42% of patients were able to complete e the questionnaire. Among these, 58% found it moderately to very much useful. For families, having an accurate prognosis influenced the planning of visits (69%), communication/closure (42%) and spiritual needs/funeral arrangements (31%). Patients identified planning of visits (10%), communication/closure (12%), and goals/accomplishments (9%) as very important. Discussing the prognosis and its impact is very helpful for the mental health professionals to have open and honest conversations with patients/families to identify, prioritize and adapt treatment to achieve goals prior to death.

Correction factors for large-scale greenhouse gas assessment from pulp and paper mill sludge landfill sites.

Assessments of greenhouse gas (GHG) emissions in managed areas are facing various challenges. A non-flow-through, non-steady-state (NFT-NSS) chamber coupled to a frame permanently inserted into the landfilled substrates is a standard method for quantifying GHG emissions in managed areas, such as pulp and paper mill sludge (PPMS) landfill sites. Frequent measurements are needed to minimize uncertainties on GHG emission factors at the landfill site scale. However, maintaining a frame inserted into the substrates for a long time period is often impossible due to landfilling management operations. Therefore, GHG measurements using NFT-NSS chambers placed directly on substrates' surface could be an interesting option. Our objectives were to determine the relationships between CO2, CH4, and N2O fluxes measured with (F + ) and without (F-) a frame inserted in the substrates' surface and to develop correction factors for fluxes measured without a frame. Measurements were made at different PPMS landfill sites in the province of Québec, Canada. Stronger GHG flux relationships were observed at the provincial (across sites) than the specific site scale: the variance in GHG fluxes from F- chambers explained up to 80 % of variance in fluxes from F + chambers. The measured CO2, CH4, and N2O fluxes in F- chambers were on average 53, 78, and 63 % lower, respectively, than those estimated by the models at provincial scale. The correction factors developed with this approach could greatly extend the number of sites where in situ GHG measurements can be done and would help refining GHG inventories at the provincial and national levels.

Development of a Multi-Criteria Decision Analysis Rating Tool to Prioritize Real-World Evidence Questions for the Canadian Real-World Evidence for Value of Cancer Drugs (CanREValue) Collaboration.

The Canadian Real-world Evidence for Value of Cancer Drugs (CanREValue) collaboration developed an MCDA rating tool to assess and prioritize potential post-market real-world evidence (RWE) questions/uncertainties emerging from public drug funding decisions in Canada. In collaboration with a group of multidisciplinary stakeholders from across Canada, the rating tool was developed following a three-step process: (1) selection of criteria to assess the importance and feasibility of an RWE question; (2) development of rating scales, application of weights and calculating aggregate scores; and (3) validation testing. An initial MCDA rating tool was developed, composed of seven criteria, divided into two groups. Group A criteria assess the importance of an RWE question by examining the (1) drug's perceived clinical benefit, (2) magnitude of uncertainty identified, and (3) relevance of the uncertainty to decision-makers. Group B criteria assess the feasibility of conducting an RWE analysis including the (1) feasibility of identifying a comparator, (2) ability to identify cases, (3) availability of comprehensive data, and (4) availability of necessary expertise and methodology. Future directions include partnering with the Canadian Agency for Drugs and Technology in Health's Provincial Advisory Group for further tool refinement and to gain insight into incorporating the tool into drug funding deliberations.

Evaluation of the Use of Anticoagulotherapy in Cancer Patients in Palliative Care Residence.

Background: Several patients admitted to palliative care residences are on anticoagulotherapy (AC). Given the risks of venous thromboembolism (VTE) and bleeding, the decision to continue or stop AC on admission remains clinically challenging. Objectives: To determine the prevalence of AC use and incidence of suspected VTE and bleeding events in palliative care patients. Methods: Retrospective cohort study including all deceased patients at a Canadian palliative care residence over two years. Results: Among the 453 patients' charts reviewed (369 with cancer), 183 (40%) were on AC at admission or <30 days earlier. Only 64 (35%) continued AC, with 78% discontinuing it during their stay. Demographic parameters were similar in the AC and non-AC groups. The incidence of suspected VTE was lower in patients pursuing AC post-admission than in those who stopped: (4.6% vs. 6.7%) and, conversely, the incidence of bleeding was higher in patients on AC: (10.8% vs. 7.6%), though these differences were not statistically significant. The risk of death in cancer patients within 72 hours of suspected VTE or bleeding event was 80% and 30%, respectively. Patients on AC had a 33% reduced risk of VTE but a 44% increased risk of bleeding. Conclusion: This study provides information on the AC use in palliative care patients. In term of survivorship, it suggests a possible advantage to continue AC to prevent a symptomatic or distressing death. Given the low incidence of events, larger powered studies will be necessary to further characterize the risks/benefits of pursuing AC in patients in palliative care residences.

Edmonton Classification System for Cancer Pain: Comparison of Pain Classification Features and Pain Intensity across Diverse Palliative Care Settings in Canada.

Background: The goal of the Edmonton Classification System for Cancer Pain (ECS-CP) is to create an international classification system for cancer pain. Previous studies reinforce the need for standardized training to ensure consistency across assessors. There is no universally accepted classification for neuropathic pain. Objectives: Our primary objective was to describe the prevalence of ECS-CP features in a diverse sample of advanced cancer patients, using assessors with standardized training. The secondary objectives were to: (1) determine the prevalence of neuropathic pain using the Neuropathic Pain Special Interest Group (NeuPSIG) criteria and (2) examine the relationship between specific predictors: ECS-CP features, age, Palliative Performance Scale, Morphine Equivalent Daily Dose (MEDD), setting, and pain intensity; and neuropathic pain. Methods: A total of 1050 adult patients with advanced cancer were recruited from 11 Canadian sites. A clinician completed the ECS-CP and NeuPSIG criteria, and collected additional information including demographics and pain intensity (now). All assessors received standardized training. Results: Of 1050 evaluable patients, 910 (87%) had cancer pain: nociceptive (n = 626; 68.8%); neuropathic (n = 227; 24.9%); incident (n = 329; 36.2%); psychological distress (n = 209; 23%); addictive behavior (n = 51; 5.6%); and normal cognition (n = 639; 70.2%). The frequencies of ECS-CP features and pain intensity scores varied across sites and settings, with more acute settings having higher frequencies of complex pain features. The overall frequency of neuropathic pain was 24.9%, ranging from 11% (hospices) to 34.2% (palliative outpatient clinic) across settings. Multivariate logistic regression analysis revealed that age <60 years, MEDD ≥19 mg, pain intensity ≥7/10, and incident pain were significant independent predictors of neuropathic pain (p < 0.05). Conclusion: The ECS-CP was able to detect salient pain features across settings. Furthermore, the frequencies of neuropathic pain utilizing the NeuPSIG criteria fits within the lower-end of literature estimates (13%-40%). Further research is warranted to validate the NeuPSIG criteria in cancer pain.