Cancer patients often suffer from refractory symptoms near death. The use of sedatives aims to relieve suffering caused by these symptoms. The practice varies broadly. The aim of this study was to evaluate the role and trends of midazolam use in cancer patients dying in a university hospital oncology ward.

The study population of this retrospective registry-based study consists of patients who died in a university hospital oncology ward in Eastern Finland in 2010–2018 (n = 639). Information about treatment decisions, midazolam use, and background factors were gathered.

During the study period, 14.7 % of the patients dying in the ward received midazolam with sedative intent prior to death. 4.7 % (n = 30) of the whole study population had continuous infusion and the rest of the midazolam use was one or multiple single doses. Documented discussion of possible palliative sedation (PS) use was found in almost one third of all patients. Out of those, eventually receiving midazolam with sedative intent, two thirds had had this discussion. The most common symptoms leading to midazolam were dyspnea, pain, and delirium. In continuous use the median midazolam infusion rate was 4.0 mg/h. The continuous infusion started median of 23.25 h and multiple single doses 19 h before death. If only one dose of midazolam was needed, it was given median of 30 minutes prior to death and the most common symptom was dyspnea. Those who received midazolam were more likely to be younger (p = 0.003) and had had a palliative outpatient clinic visit (p = 0.045).

This is the first study to report the trends and practices of midazolam use for refractory symptoms in Finland. Midazolam was used for approximately every 7th dying cancer patient. Applying midazolam was supported by a history of palliative clinic visits and younger age.

Continuous deep sedation (CDS) is a way to reduce conscious experience of symptoms of severe suffering in terminally ill cancer patients. However, there is wide variation in the frequency of its reported. So we conducted a retrospective analysis to assess the prevalence and features of CDS in our palliative care unit (PCU).

We performed a systemic retrospective analysis of the medical and nursing records of all 1581 cancer patients who died at the PCU at Higashi Sapporo Hospital between April 2005 and August 2011. Continuous deep sedation can only be administered safely and appropriately when a multidisciplinary team is involved in the decision-making process. Prior to administration of CDS, a multidisciplinary team conference (MDTC) was held with respect to all the patients considered for CDS by an attending physician. The main outcome measures were the frequency and characteristics of CDS (patient background, all target symptoms, medications used for sedation, duration, family's satisfaction, and distress). We mailed anonymous questionnaires to bereaved families in August 2011.

Of 1581 deceased patients, 22 (1.39%) had received CDS. Physical exhaustion 8 (36.4%), dyspnea 7 (31.8%), and pain 5 (22.7%) were the most frequently mentioned indications. Continuous deep sedation had a duration of less than 1 week in 17 (77.3%). Six patients (0.38%) did not meet the appropriate criteria for CDS according to the MDTC and so did not receive it. Although bereaved families were generally comfortable with the practice of CDS, some expressed a high level of emotional distress.

Our results indicate that the prevalence of CDS will be decreased when it is carried out solely for appropriate indications. Continuity of teamwork, good coordination, exchange of information, and communication between the various care providers are essential. A lack of any of these may lead to inadequate assessment, information discrepancies, and unrest.