[Analysis of palliative sedation in hospitalised elderly patients: Effectiveness of a protocol].

OBJECTIVE

To measure changes in the practice of palliative sedation during agony in hospitalised elderly patients before and after the implementation of a palliative sedation protocol.

MATERIAL AND METHODS

A retrospective before-after study was performed in hospitalised patients over 65 years old who received midazolam during hospital admission and died in the hospital in two 3-month periods, before and after the implementation of the protocol. Non-sedative uses of midazolam and patients in intensive care were excluded. Patient and admission characteristics, the consent process, withdrawal of life-sustaining treatments, and the sedation process (refractory symptom treated, drug doses, assessment and use of other drugs) were recorded. Association was analysed using the Chi(2) and Student t tests.

RESULTS

A total of 143 patients were included, with no significant differences between groups in demographic characteristics or symptoms. Do not resuscitate (DNR) orders were recorded in approximately 70% of the subjects of each group, and informed consent for sedation was recorded in 91% before vs. 84% after the protocol. Induction and maintenance doses of midazolam followed protocol recommendations in 1.3% before vs 10.4% after the protocol was implemented (P=.02) and adequate rescue doses were used in 1.3% vs 11.9% respectively (P=.01). Midazolam doses were significantly lower (9.86mg vs 18.67mg, P<.001) when the protocol was used than when it was not used. Ramsay sedation score was used in 8% vs. 12% and the Palliative Care Team was involved in 35.5% and 16.4% of the cases (P=.008) before and after the protocol, respectively.

CONCLUSIONS

Use of midazolam slightly improved after the implementation of a hospital protocol on palliative sedation. The percentage of adequate sedations and the general process of sedation were mostly unchanged by the protocol. More education and further assessment is needed to gauge the effect of these measures in the future.