Two studies (n=263) met all of the inclusion criteria and were included in the review: one was a randomised controlled trial (RCT; n=99) and one was a prospective cohort study with a comparison group (n=164). Two other studies (n=70) did not formally meet the inclusion criteria as they were not controlled: one was a description of a care programme and one was the description of the adoption of guidelines. Data on these two studies were extracted but were not included in the synthesis.
In the RCT, the treatment groups were similar with respect to baseline disease severity and demographic characteristics, all patients entered were accounted for, and the outcome assessor was blinded to the treatment groups. In the cohort study, the treatment groups differed at baseline and researchers were not blinded to the treatment group.
The RCT found that, compared with non-palliative procedures, a palliative care intervention in an acute general hospital setting had no influence on mortality, site of discharge, length of hospital stay, number of readmissions, use of non-palliative procedures, do-not-resuscitate orders and antibiotic use. Patients in the intervention group were significantly more likely to have an overall palliative care plan (P=0.008).
The cohort study found that TLTC patients received significantly more analgesia, antibiotics, intravenous therapy, hypnotics and anxiolytics than DSCU patients (P<0.05). DSCU patients also experienced significantly less discomfort, and fewer were transferred to acute medical settings than TLTC patients. DSCU patients had a higher mortality rate than TLTC patients (hazard ratio 2.2).