Background Intravenous morphine is the preferred drug for the treatment of moderate to severe pain by paramedics. Nausea and vomiting are believed to be frequent side-effects and routine co-administration of metoclopramide is common. In the absence of pre-hospital data to support this practice, we sought to determine the incidence of peri-opiate nausea and vomiting in an ambulance service which does not administer anti-emetics.

Methods This prospective observational study is currently assessing the incidence of emesis in 400 patients attended by the North East Ambulance Service, aged above 17 years and receiving morphine, using a patient-scored Nausea and Vomiting Score (NVS: 0=no nausea or vomiting, 1=slight nausea, 2=moderate nausea, 3=severe nausea, 4=vomited once, 5=vomited twice or more).

Results To date 145 patients have been recruited. Median NVS before morphine was 0 (range 0 to 6, inter-quartile range (IQR) 0 to 1): 54/141 (38%) of patients had some degree of nausea or vomiting. Median NVS on hospital arrival (after morphine) was 0 (range of 0 to 6, IQR 0 to 1): 54/130 (42%) patients had some degree of nausea or vomiting. The differences pre- vs. post-morphine in median NVS (p=0.98) and proportion of patients suffering nausea and vomiting are not statistically significant (p=0.98 and p=0.54 respectively). There were no significant correlations between pre-morphine pain score and pre-morphine NVS; post-morphine pain score and post-morphine NVS; pre-morphine NVS and total morphine dose; and post-morphine NVS and total morphine dose (Spearman's rank correlation 0.09, p=0.274; 0.07, p=0.44; 0.10, p=0.25; and 0.10, p=0.24 respectively).

Conclusion and recommendations To date this study has found no evidence that pre-hospital administration of morphine is associated with an increased incidence or severity of nausea and vomiting and therefore does not appear to support the routine co-administration of metoclopramide.