Effectiveness of integration of palliative medicine specialist services into the intensive care unit of a community teaching hospital.

BACKGROUND: Collaboration between palliative medicine and the intensive care unit (ICU) represents best practice and offers important benefits. However, achieving effective collaboration between these two specialties can be challenging. OBJECTIVE: Assess effectiveness of integrating palliative medicine specialists in the ICU. DESIGN: Retrospective chart review. RESULTS: Of 201 patients who qualified for palliative consultation using a palliative screening tool, 92 were referred and 109 were not referred for palliative medicine consultation. The number of screening criteria met was similar between the two groups. Palliative medicine consult volume increased significantly compared with preintegration (7.7±3.4 versus 4.4±2.8 consults per month, p=0.04). No significant difference in hospital mortality was found between the referred and unreferred groups (32/92 [35%] versus 26/109 [24%], p=0.09). ICU length of stay was significantly shorter in the referred group (7 versus 11 days, p<0.001). Referred patients were more frequently enrolled in hospice compared with unreferred patients (32/92 [37%] versus 3/109 [3%], p<0.001). ICU physicians referred patients significantly more often for dementia and ventilator withdrawal (13/16, p=0.003; 24/29, p<0.001, respectively) and significantly less often for ICU stay longer than 10 days (21 versus 49, p=0.001). CONCLUSIONS: Integrating palliative medicine specialists into intensive care was associated with a significant increase in use of palliative medicine services and a significant decrease in ICU length of stay for referred patients without a significant increase in mortality. The screening tool effectively identified patients at high risk of death. Given the high mortality rate of the unreferred patients, the criteria could be more widely adopted by ICU physicians to consider expanding palliative medicine referrals.

Impact of standardized palliative care order set on end-of-life care in a community teaching hospital.

BACKGROUND: We evaluated use of medications and interventions in patients receiving a new standardized palliative care order set (PCOS) compared with patients receiving no palliative care orders and those with an order for comfort measures only (CMO), the largely ineffective method used at our institution before implementation of the order set. METHODS AND RESULTS: We conducted a retrospective chart review of all patients who died at our community teaching hospital between November 2006, 8 months after PCOS implementation, and May 2007. Of 106 patients in the study group, 31 (29%) were treated using the PCOS, 6 (6%) received the CMO order, and 69 (65%) did not receive either. Patients in the PCOS group had significantly more orders for every palliative medication (p ≤ 0.05). Opioids and anxiolytics were made available to every patient in the PCOS group. Most PCOS patients received orders for antiemetic, antipsychotic, antisecretion, and laxative medication during the end-of-life period. No CMO patients and few patients in the no palliative care orders group received orders for these medications. The PCOS and CMO group similarly limited nonpalliative interventions, whereas the nonpalliative group had relatively high use of these interventions until death. CONCLUSION: The palliative care order set implemented at our community teaching hospital significantly improved adherence to accepted palliative care treatment principles for patients at the end of life.

Impact of writing "comfort measures only" orders in a community teaching hospital.

OBJECTIVES: We compared end-of-life care for patients with and without orders for "comfort measures only" (CMO) and evaluated whether standards for palliative medicine were met. DESIGN: We retrospectively reviewed the charts of 40 patients, 20 with and 20 without a CMO order, who died in our community teaching hospital from October 2003 through September 2004. Appropriate use of medications and interventions and documentation of symptoms and family consultation were examined. RESULTS: There were no significant demographic differences between the groups. Family meeting (18/20 versus 7/20, p < 0.001) and do-not-resuscitate/do-not-intubate (DNR/DNI) code status (20/20 versus 8/20; p < 0.001) occurred significantly more frequently in the CMO group. There were no other significant differences in prescribing medications or treatments. Notation of symptoms common at the end of life did not differ between the groups except for nausea/vomiting, which was more common in the non-CMO group. In the CMO group, opioids were significantly more available and antibiotics and routine laboratory work were significantly less common after the CMO order, but no other differences were observed. CONCLUSION: CMO orders alone were insufficient for redirecting changes in care at the end-of-life.