Patient Preferences for Discussing Life Expectancy: a Systematic Review.

BACKGROUND: Discussing life expectancy helps inform decisions related to preventive medication, screening, and personal care planning. Our aim was to systematically review the literature on patient preferences for discussing life expectancy and to identify predictors for these preferences. METHODS: We searched PubMed, Cochrane Library, Embase, MEDLINE, PsycInfo, and gray literature from inception until 17 February 2021. Two authors screened titles/abstracts and full texts, and extracted data and one author assessed quality. The outcome of interest was the proportion of patients willing to discuss life expectancy. We reported descriptive statistics, performed a narrative synthesis, and explored sub-groups of patients according to patient characteristics. RESULTS: A total of 41 studies with an accumulated population of 27,570 participants were included, comprising quantitative survey/questionnaire studies (n=27) and qualitative interview studies (n=14). Willingness to discuss life expectancy ranged from 19 to 100% (median 61%, interquartile range (IQR) 50-73) across studies, with the majority (77%) reporting more than half of subjects willing to discuss. There was considerable heterogeneity in willingness to discuss life expectancy, even between studies from patients with similar ages, diseases, and cultural profiles. The highest variability in willingness to discuss was found among patients with cancer (range 19-100%, median 61%, IQR 51-81) and patients aged 50-64 years (range 19-97%, median 61%, IQR 45-87). This made it impossible to determine predictors for willingness to discuss life expectancy. DISCUSSION: Most patients are willing to discuss life expectancy; however, a substantial proportion is not. Heterogeneity and variability in preferences make it challenging to identify clear predictors of willingness to discuss. Variability in preferences may to some extent be influenced by age, disease, and cultural differences. These findings highlight the individual and complex nature in which patients approach this topic and stress the importance of clinicians considering eliciting patient's individual preferences when initiating discussions about life expectancy.

Mathematical arterialization of venous blood in emergency medicine patients.

OBJECTIVES: Arterial punctures represent a painful and unpleasant experience. Acid-base and oxygenation status can be assessed from peripheral venous blood, but agreement with arterial values is not always clinically acceptable. This study evaluates a method for mathematically transforming peripheral venous values into arterial values in emergency medicine patients. METHODS: Paired arterial and peripheral venous samples were analysed in groups A (47 patients) and B (101 patients), corresponding to the clinical need for arterial blood sampling (A) and without (B). Venous values were input into the mathematical arterialization method and the values of arterial pH, PCO2 and PO2 were calculated and compared with the measured values. RESULTS: The calculated and measured arterial pH and PCO2 values correlated well with the correlation coefficients (r ) of group A, pH 0.94, PCO2 0.97; group B, pH 0.87, PCO2 0.83; and Bland-Altman limits of agreement well within the limits of acceptable laboratory and clinical performance. The calculated values of arterial PO2 followed a set of predefined rules relating calculated and measured PO2 levels in all cases. The method represents an improvement on the use of venous blood alone where the correlation coefficients were as follows: group A, pH 0.85, PCO2 0.88; group B, pH 0.79, PCO2 0.59; and limits of agreement for PCO2 at the border of (group A) or beyond (group B) acceptable clinical limits. CONCLUSION: Application of the mathematical arterialization method may reduce the pain associated with assessment of acid-base and oxygenation status, maximize the information obtained from peripheral venous blood and allow venous measurements to be presented as more commonly interpreted arterial values.