ARE PRESCRIBED INFUSIONS RUNNING AS INTENDED?
log is required. A quality timestamp would provide more accurate information about the order of device events and the precise duration elapsed between these events, especially when multiple devices connected to the same patient are being analysed. For the logs analysed in this study, the timestamp of an event was precise to the minute and, as we found during our analyses, multiple events could happen within the same minute. We would recommend that log timestamps be precise down to the millisecond and that all user keystrokes on infusion devices be logged. (Note that precise logging of time is a different matter from accurate logging; the clocks on the pumps may be inaccurate compared to actual GMT, but their interval measurements need to be accurate as well as precise.)
In general, we recommend regular analysis of event and error logs to provide feedback on the use of devices for different stakeholders, as identifi ed below.
Pharmacy Pharmacies dispensing medications could use log analysis to see the absolute time typically taken to deliver time- critical medications. Such medication might be better prescribed in smaller bags to minimise the possibility of expiry during the course of an infusion interruption.
Some of the infusion interruptions observed in the logs seem possibly to be the result of the pump’s user interface design. For instance, occlusion (pause pumping on the Graseby 500 which will lead to an occlusion if not dealt with in 10 seconds), is seen to be one of the most frequent events that cause an infusion to be interrupted. The high frequency of occlusion events could be a result of the poor accessibility for the user of the pressure-level information on the pump. Users typically do not adjust the pump’s occlusion pressure level when the rate changes from high to low or vice versa (Lee et al, 2012). Further studies are required to test this hypothesis.
Equipment management and maintenance It is common practice for pumps used in hospitals to be managed effi ciently using a library, which allows a hospital to minimise costs of purchasing and maintaining the pumps they use. The average number of infusions per department, the average number of infusions per pump, and the average length of each infusion can be used to plan a better maintenance program for the pumps.
Regular analyses of logs may show deterioration in battery life. According to Smiths Medical Ltd (2008), the
20 CareFusion Supplement BJN July 2013
› We recommend regular analysis of event and error logs to provide feedback ‹
Graseby 500 pump should sound an alarm when it has 30 minutes of battery life left. Oladimeji et al (2011) and Lee et al (2012) both show in their analyses of logs that a signifi cant number of low-battery alarms during infusions lead on to dead battery events. Some batteries are completely depleting approximately 2 minutes after low-battery alarms, suggesting battery capacity problems that were going unnoticed. The MHRA also cited several incidents involving battery failures (MHRA, 2005). These and other hardware-related issues would be noticed if logs were analysed regularly, leading to appropriate maintenance action that could consequently prevent interruption to therapy that may lead to patient harm.
Training Understanding how infusions actually run in practice would be of benefi t for those who train staff to use infusion devices. Ineffi cient set-up procedures can be inferred from the logs and corrected during staff training. Workarounds may also be inferred from log analysis and these can be optimised or reinforced in training, particularly if they address device design fl aws.
Studying the log data fi les provides an insight into the way infusion pumps are used in the clinical setting. This study could be replicated in other hospitals, expanding the data available and allowing comparison between hospitals. Common features would be identifi ed as a consequence of this comparison and general guidelines or recommendations for the design, usage and acquisition of infusion pumps could be created.
Funded as part of the CHI+MED: Multidisciplinary Computer-Human Interaction research for the design and safe use of interactive medical devices project, supported by EPSRC [Grant Number EP/G059063/1].
References Barber N, Taxis K (2004) Incidence and severity of intravenous drug errors in a German hospital. Eur J Clin Pharmacol 59(11): 815–7
Lee PT, Thompson F, Thimbleby H (2012) Analysis of infusion pump error logs and their signifi cance for health care. Br J Nurs 21(8): S12–20
| Page 2
| Page 3
| Page 4
| Page 5
| Page 6
| Page 7
| Page 8
| Page 9
| Page 10
| Page 11
| Page 12
| Page 13
| Page 14
| Page 15
| Page 16
| Page 17
| Page 18
| Page 19
| Page 20
| Page 21
| Page 22
| Page 23
| Page 24
| Page 25
| Page 26
| Page 27
| Page 28