Introduction Infusion pumps play a key role in the life of patients on home parenteral nutrition. In addition to ensuring the accurate delivery of fluid, pre-set alarms alert patients to potential infusion or catheter related problems. Pumps have typically not permitted analysis of this information meaning that the nature and frequency of alarms is not known. Industry collaboration allowed the development of a unique system which would record details of the type and time of all pump alarms and send a notification email via a secure server to named healthcare professionals caring for the patient. The aim of this study was to identify the characteristics and rate of infusion alarms in a group of patients receiving home parenteral nutrition.

Method 20 patients consented to have their infusions monitored. Any non intentional pump alarm (downstream occlusion, upstream occlusion, air in line or “other”, such as low battery, door open) generated a real time email alert outlining the time and nature of alarm to the secure email address of the lead investigator.

Results 20 patients were monitored for 3 months. Patients infused an average of 5.3 ± 1.7 (2–7) nights per week. During the assessment period 922 email alerts were received, mean 46 ± 41.9 per patient, range 3–160. Adjusting for the number of nights patients infused this equated to a mean of 0.6 ± 0.6 alarms per infusion night. Patients experienced alarms on an average of 27 ± 0.21% of their infusion nights, range 4–76%. Only 1 patient (5%) experienced alarms on more than 75% of their infusion nights. 4 patients (20%) had alarms on less than 5% of their infusion nights. All patients had nights with no alarms. There was considerable variation in the largest number of alarms experienced per single infusion, mean 13.5 ± 19.8, range 1–93. There were significantly more downstream occlusions (n = 622) vs. upstream occlusions (n = 223), air in line (n = 37) or “other” (n = 35) alarms, p = 0.05. Of the “other” alarms, the most frequent was low battery (n = 25), followed by door open (n = 6) and an internal error code requiring maintenance (n = 4). The 2 patients with the highest occurrence of downstream alarms were the only patients that had ports.

Conclusion This is the first time that the type and frequency of pump alarms in patients on home parenteral nutrition has been quantified. It demonstrates variation in the number of alarms patients receive, with some patients experiencing very few alarms. The technology provides valuable real time monitoring which could permit early intervention in identifying pump or infusion related problems.

Disclosure of interest M. Small: None Declared, M. Kallis Paid Instructor for: Micrel Medical Devices, A. Tsoukalis Employee of: Micrel Medical Devices, J. Cascailh Employee of: Micrel Medical Devices, S. Gabe: None Declared.