Palliative Care Physicians' Practice in the Titration of Parenteral Opioids for Dyspnea in Terminally Ill Cancer Patients: A Nationwide Survey

We conducted a cross-sectional survey among palliative care physicians in Japan between September and October 2018. The single inclusion criterion was that all participants were palliative care physicians certified by the Japanese Society for Palliative Medicine. We identified a total of 536 certified physicians from the Society's online registry and sent questionnaires to 268 randomly selected physicians.

Because of the lack of an existing specific questionnaire, we developed an ad hoc questionnaire on the basis of literature reviews,^{,  ,  ,  ,  ,}  pilot testing, and discussion among the authors. The questionnaire included several components of physician-reported practice regarding parenteral opioids for dyspnea at rest in terminally ill cancer patients (defined as those with ECOG performance status = 4, with an estimated prognosis one or two weeks). The first component examined physician-reported practice in a hypothetical scenario where a terminally ill cancer patient continued to suffer severe dyspnea despite titration of continuous administration of parenteral morphine. Further titration would be considered clinically controversial because uncomfortable side effects have developed (no respiratory depression). The response options included the following: 1) adding a small amount of midazolam by continuous administration (e.g., ≤10 mg/day), 2) adding benzodiazepines other than midazolam, 3) titrating morphine further, and 4) switching to another opioid. These were rated on a five-point Likert scale (1 = rarely – 5 = very frequently). The participants were also asked about the starting dose of midazolam for dyspnea which is to be added to parenteral morphine in their usual practice. The second component investigated the participants' specific practice regarding titration of parenteral morphine, such as the timing of reevaluation after the initiation or dose adjustment, intervals of dose adjustment (either increase due to suboptimal relief or decrease due to side effects), and rescue doses and intervals. Third, we investigated the participants' opinion about the presence of the upper limit of morphine for dyspnea above which further titration is deemed ineffective (yes/no), as well as physician-perceived upper limits for both opioid-naïve (mg/day) and opioid-tolerant patients (% increase from the baseline).

Descriptive statistic analyses were conducted using SPSS software (version 25; IBM, Tokyo, Japan). Assuming that 50%–70% of participants would frequently or very frequently perform a certain practice, 143–171 subjects would be sufficient to calculate an accuracy to within a 15% width with 95% CIs. This study was approved by the Institutional Review Board of Konan Hospital.

A total of 189 physicians responded (response rate, 71%).

To our knowledge, this is the first nationwide survey among palliative care physicians to investigate their detailed practice in the use of parenteral opioids for dyspnea in terminally ill cancer patients.

The first important finding was that the addition of a small amount of midazolam by continuous administration was the most prevalent practice followed by the regular administration of benzodiazepines other than midazolam when further opioid titration was considered clinically controversial. This is largely consistent with recommendations from previous guidelines,^,  although the median dose was lower than that previously reported in clinical trials and guidelines.^,  Future studies should explore if the addition of midazolam is truly effective for dyspnea, and if so, the most effective dose.

The second important finding was that approximately a third of physicians agreed that there was an upper limit of opioids for dyspnea in both opioid-naïve and opioid-tolerant patients. The perceived upper limits were relatively low, which was consistent with a previous study in chronic dyspnea which revealed that most dyspnea could be managed with a lower dose of oral morphine (10–30 mg/day). However, the upper limit of opioids for dyspnea in terminally ill cancer patients has not been explored in clinical trials or empirical studies. These findings contrast sharply with pain management in which an upper limit of opioids is widely considered nonexistent. A dose-effect relationship in opioids for dyspnea in terminally ill patients should be investigated in future studies.

The strengths of this study were sampling from the Society's registry of certified palliative care physicians and a relatively high response rate, which ensure generalizability. The limitations included the fact that the questionnaires were not validated and that physician-reported practice might not fully reflect their actual practice. In addition, to maximize a response rate by minimizing participants' burden, we did not obtain detailed demographic data of the participants and achieved the high response rate (71%). Our aim was to explore the overall practice patterns of the board-certified palliative care physicians rather than differences in practice among various subgroups of physicians.

On titrating parenteral opioids for severe dyspnea in terminally ill cancer patients, it was reported that most palliative care physicians would add a small amount of midazolam by continuous administration when further opioid titration is considered clinically controversial. Future research is warranted to establish a standardized parenteral opioid treatment for dyspnea in terminally ill cancer patients.

The authors declare no conflict of interest.

This study was in part supported by Grant-in-Aid for Young Scientists of the Ministry of Education, Culture, Sports, Science and Technology of Japan, and Japan Society for the Promotion of Science KAKENHI grant number JP16K15418. The authors would like to thank Drs. Keiko Tanaka and Tatsuya Morita for their valuable advice on their study.