Related Projects

full title

Timing and collaboration in the Neonatal Intensive Care Unit


Timing, Collaboration, Neonatal Intensive Care, Cognitive Task Analysis

This study investigated how timing and collaboration issues in human-machine interaction (HMI) affect dependability in neonatal intensive care. An expert system, FLORENCE (Fuzzy LOgic for REspiratory Neonatal Care Expert), is being developed at St James’s University Hospital (Leeds) to help senior house officers (junior doctors) make decisions about changes to the ventilators that are regularly used in treating premature babies. A case study of the current socio-technical system that implements the practice of neonatal intensive care at SJUH identified several timing and collaboration issues related to the dependability of the system. These issues were used to generate requirements for the design and use of FLORENCE that should help to ensure that the dependability of the new system (including FLORENCE) should be at least equal to the dependability of the system without FLORENCE.

A Cognitive Task Analysis of the NICU uncovered examples of the following timing issues:

• Sequences, e.g. check baby’s airways, breathing then circulation (ABC);
• Delays, e.g., wait to check alarm is real before responding;
• Delayed Feedback, e.g., after changing the ventilator settings there is a 2-3 minute delay before change in blood gases become apparent
• Trends, e.g., improving state of blood gases should start to be visible 2-3 minutes after changes are made to ventilator settings;
• Temporal validity, e.g., manual blood gas data is only valid for 20-30 minutes;
• Deadlines, e.g., self-imposed deadlines for detecting improvement in baby’s condition;
• Pacing, e.g., taking observations on the hour boundary;
• Other, e.g., the inspiration and expiration time can be controlled for breaths delivered by the ventilator
The collaboration issues that were identified were:
• Organisational structure, e.g., the hierarchy inherent in communication and decision making
• Formal verbal communication, e.g., the daily ward round which implements the formal handover between the night shift and the day shift;
• Informal verbal communication, e.g., separate informal shift handovers for the nurses and the SHOs;
• Formal non-verbal communication, e.g., the extensive use of records in the NICU by all staff.

The implications of these findings can be summarised in two key aspects that should help to ensure that FLORENCE does not make the system of neonatal intensive care less dependable. The first lies in the education of the front line carers. During the SHOs induction training, the role and purpose of FLORENCE needs to be very clearly explained so that they know how FLORENCE integrates with the other aspects of their work. The second is the monitoring of the use of FLORENCE. As part of the checking of the progress of the SHOs, senior staff will need to monitor the usage of FLORENCE to make sure that front line carers do not become too trusting or over-reliant on FLORENCE. This will necessarily include checking how much SHOs are learning from communicating and collaborating with other staff. The process of monitoring the use of FLORENCE will therefore need to be an ongoing one.

Author: Gordon Baxter


Timeliness theme

Using Cognitive Task Analysis to assess dependability in the Neonatal Unit


Baxter, G.D., Filipe, J-K., Miguel, A. & Tan, K. (2005). The effects of timing and collaboration on dependability in the neonatal intensive care unit. In. F.Redmill & T. Anderson (Eds.), Constituents of modern system-safety thinking: Proceedings of the thirteenth safety-critical systems symposium (pp. 195-210). Berlin, Germany: Springer Verlag.


Page Maintainer: Credits      Project Members only Last Modified: 30 May, 2006