Context The ability to predict survival accurately has implications in clinical decision making. Objective We determined the association of phase angle obtained from multi-frequency bioelectric impedance analysis (MF-BIA) with overall survival in patients with advanced cancer. Methods We included consecutive patients with advanced cancer who had an outpatient palliative care consultation. MF-BIA assessed phase angle at three different frequencies (5 kHz/50 kHz/250 kHz) on each hemibody (right/left). Survival analysis was conducted using the Kaplan Meier method, log rank test and multivariate Cox regression analysis. Results Among 366 patients, the median overall survival was 250 days (95% confidence interval 191–303 days). The mean phase angle for 5 kHz, 50 kHz and 250 kHz were 2.2°, 4.4°, 4.2° on the right, and 2.0°, 4.2° and 4.1° on the left, respectively. For all 6 phase angles, a lower value was significantly associated with a poorer overall survival (P<0.001). After adjusting for cancer type, performance status, weight loss and inflammatory markers, phase angle remained independently associated with overall survival (hazard ratio 0.85 per degree increase, 95% confidence interval 0.72–0.99; P=0.048). Conclusion Phase angle represents a novel objective prognostic factor in outpatient palliative cancer care setting, regardless of frequency and body sides.

Background. It is unclear how oncologists' attitudes toward end-of-life (EOL) care affect the delivery of care. The present study examined the association between oncologists' EOL care attitudes and (a) timely specialist palliative care referral, (b) provision of supportive care, and (c) EOL cancer treatment decisions. Methods. We randomly surveyed 240 oncology specialists at our tertiary care cancer center to assess their attitudes toward EOL care using a score derived from the Jackson et al. qualitative conceptual framework (0 5 uncomfortable and 8 5 highly comfortable with EOL care). We determined the association between this score and clinicians' report of specialist palliative care referral, provision of supportive care, and EOL cancer treatment decisions. Results. Of the 182 respondents (response rate of 76%), the median composite EOL care score was 6 (interquartile range,

144 Background: There is limited literature regarding outpatient palliative care, and even more limited literature describing factors associated with unscheduled visits. Our aim was to identify clinical characteristics of patients who walk-in (present unscheduled) to the outpatient SCC with the hypothesize that the patients who present for unscheduled visits have more severe symptom distress as compared to those patients who present for scheduled appointments. Methods: We compared 183 patients with unscheduled consults (CO) vs a random sample of 361 patients with scheduled CO and 159 patients with unscheduled follow-up (FU) visits vs a random sample of 318 patients with scheduled FU visits. Results: Among 544 total CO visits, unscheduled visits had worse median Edmonton Symptom Assessment Scale (ESAS) symptoms (on a scale from 0 to 10): pain (6.5 vs 4.7, p < .0001), fatigue (p = .002), nausea (p = .017), depression (p = .003), anxiety (p = 0.02), sleep (p = .0002), and overall feeling of well-being (p = 0.0009). There was no statistical difference in shortness of breath, financial distress, nor spiritual pain. Daily opioid dose (MEDD in mg) was significantly higher in unscheduled CO visits (119.7 vs 62.9, p = .0004). Among 344 total FU visits, unscheduled visits had worse median ESAS symptoms: pain (5.7 vs 4.2, p = .0001), fatigue (p = .0006), depression (p = .019), anxiety (p = .014) and higher MEDD (111.3 vs 73.6, p = .0009). There was no difference in type of insurance coverage and better ECOG (p = .015) in unscheduled vs scheduled CO visits. Unscheduled CO visits were more likely to be from outside the Houston area (161/361, 45% vs. 40/133, 30%, p < 0.0001). Conclusions: Patients who are either referred as unscheduled CO visits or who present as unscheduled FU visits have higher levels of physical and psychosocial distress and higher opioid dose. Outpatient palliative care centers need to provide opportunity for walk-in referrals for timely management of these issues.

240 Background: There are potential severe effects when patients taking opioids receive other psychoactive medications. However, such combinations are sometimes necessary in palliative care. The purpose of this study was to determine the frequency of concomitant use of opioids + psychoactive medications in cancer patients referred to our outpatient palliative care center. Methods: Retrospective data obtained from consecutive consults was analyzed to determine the frequency of patients on opioids alone versus concomitant opioids + psychoactive medications at first presentation to our clinic. Association of type of medication with demographics and baseline characteristics was evaluated by Wilcoxon rank sum test for continuous variables and Chi-square (Fisher's exact) test for categorical variables. Results: Among 541 consecutive consult visits, 365 (67%) patients were taking opioids at the time of referral to our clinic: 209 (57%) were on opioids alone while 156 (43%) were on concomitant opioids + psychoactive medications [69 (44%) were on Opioid + Benzodiazepine, 46 (30%) were Opioid + Antidepressants, 41(26%) were on both). Patients in the concomitant groups were on higher Morphine Equivalent Daily Dose (MEDD, p = 0.007), had higher Edmonton Symptom Assessment Scores (ESAS) for pain (p = 0.017), anxiety (p < 0.001), depression (p < 0.001) and spiritual pain (p = 0.03). Conclusions: A large proportion (156, 43%) of cancer patients referred to outpatient palliative care was on concomitant opioids + psychoactive medications. These patients were on higher doses of opioids with higher levels of pain and psycho-social distress at the time of first presentation. Further studies are required to better understand the clinical implications of concomitant use of opioids + psychoactive medications in such patients.

No abstract