Designing out medical error

Designing Out Medical Error, DOME, was a three year, EPSRC-funded, multidisciplinary project aimed at designing safer healthcare equipment, processes and devices. The project brought together clinicians, designers, psychologists, human factors and business expertise to develop solutions to potential errors in and around the hospital bed-space. By taking a systems approach to design, the project identified causes common to a range of failures in the highest risk processes. The project has produced a range of designs including the CareCentre™, which is now in clinical trials.

There are around 100,000 reports of patient safety incidents per month to the National Reporting and Learning Service from NHS Trusts in England and Wales. Other sources of estimated harm place the rate of in-hospital adverse events (unintended injury or complication) at almost 1 in 10 admissions.

Whilst ergonomics/human factors is now a constituent of safety management and systems design in many industries, healthcare is one area in which it doesn’t play a large role. Yet the complexity and distributed nature of healthcare means there are extensive opportunities for human factors to contribute to improvements. A single patient journey can cross boundaries between the primary, secondary and tertiary healthcare sectors, and care is often delivered by distributed teams working in emergency situations within unfamiliar and multifunction workspaces. Medical devices are a component of virtually all healthcare processes from bedside lockers through to life support systems, but usability issues are often under recognised or reported.

The objectives of the DOME study were:

  • To develop a multidisciplinary approach to designing for patient safety that would provide long term engagement and potential for future design collaborations.
  • To map, analyse and prioritise the hazards in a surgical ward.
  • To develop design solutions using a systems approach and co-design methods.

DOME took a systems approach to the design of the healthcare processes, equipment, environment and information used in the bed-space of a typical surgical ward. Applying this approach to such a common workspace means that the methods and solutions will be transferable to many other clinical specialties and settings. Rather than focusing on a particular type of error, incident or activity, the project took a holistic view of the activities that take place around the bedside. This approach allowed the systems’ influences on the safety of all aspects of care to be considered in parallel. A collaborative approach saw designers and clinicians involved in joint observations, data collection and design activities, and facilitated knowledge and skill transfer between disciplines.

A mixed-methods approach was adopted, utilising methods from psychology, human factors, social science, operational management and design. This approach included observations, interviews, surveys, shadowing, mapping of healthcare processes, Failure Modes and Effects Analysis, risk identification and prioritisation, focus groups, causal analysis, generation of design briefs, concept generation, brain storming, co-design, simulation studies and clinical trials and evaluations.

A work analysis was conducted based on 70 hours of observation on five general surgery wards at three hospitals during the day, night and weekend. The work analysis identified 14 top-level healthcare processes around the bedside, each with numerous sub-processes, such as:

  • Hand hygiene e.g. alcohol gel, soap, water
  • Vital signs monitoring e.g. blood pressure, pulse, respiratory rate
  • Ward round e.g. history taking, writing notes/forms
  • Medication delivery e.g dispensing medication, management of IV drips and stands.

Given the large number of activities observed, the processes were risk assessed by healthcare workers, patients and visitors using subjective rating scales. The highest risk processes were identified as hand hygiene, vital signs monitoring, isolation of infection, medication delivery and handover of information. A Healthcare Failure Modes and Effects Analysis was then used to identify how each of the surgical ward healthcare processes could fail.

Nearly 200 potential failure modes were identified in just these five processes. Further analysis of the top 60 failure modes identified the contributory factors. Design was cited as a contributor to the high risk failures in all five processes. A lack of reminders, and poor monitoring of staff performance and feedback were also common, together with a lack of standardisation, issues with leadership, clear team roles and responsibilities, education, training and patient safety not being a priority.

This research was then translated into a set of meaningful design briefs which captured the research findings providing realistic boundaries for design work and they inspired a breadth of ideas. The decision was taken to base the briefs on the broad understanding of the failures in each of the five processes rather than rooting them in certain, specific failures. For instance, rather than focusing on a specific failure such as the step “decide to wear gloves and apron”, the briefs addressed the failures and the causes across the whole process. Also, by considering each process simultaneously, the design work addressed potential failures in the system of interlinked processes which make up the typical activity around the bedside. Five interlinked briefs were therefore formed around the processes of hand hygiene, isolation of infection, vital signs monitoring, medication delivery and handover of information. These briefs were validated by a process expert in each case to ensure they captured the findings of the research.

An iterative design process was followed where ideas were continuously presented to healthcare workers and patients for critical input. One of these, the CareCentre™, was developed primarily in response to the isolation of infection brief, although it also addresses many of the other briefs, demonstrating the merit of considering multiple processes and design briefs simultaneously. Shadowing nurses trying to follow correct protocols revealed that much time was wasted searching for gloves and aprons which were often located far from the bedside.

These observations were extended to include the use of other equipment for common bedside processes. It was found that the medication locker was often inaccessible (located on the wall, often with a patient obstructing), gloves and aprons were situated away from the bedside, there was no flat surface for reviewing or writing documents, cleaning wipes were not within easy reach (again, located on the wall), and the hand gel at the foot of the bed was difficult to access from the bedside.

The concept of rationalising all this equipment into a ‘one-stop-shop’ met with user approval, and through a series of feedback sessions with front line staff, the list of contents of this all-in-one unit was defined, as well as its position at the end of the bed.

The first prototype was produced and taken to over 120 staff for review. This featured a flat surface for writing documents, a medication locker, hand gel, cleaning wipes, aprons and gloves, and a folder holder to contain the patient’s charts. The concept was designed to hook over the end of the bed. Following further iterations and user feedback, the CareCentre™ is now in manufacture and is in clinical trials.
Other outputs from the study include signage for hand hygiene, a re-designed vital signs trolley with retractable leads, a mobile phone application for recording respiratory rate, a medication dispenser that records missed medication, and design requirements for a handover space.

By Beverley Norris, Jonathan West & Oliver Anderson