Using Safety-II and resilient healthcare principles to learn from Never Events

[Note. This is an update & repost of an older post but now in article format]

A very interesting read. This used a secondary analysis (qualitative and quantitative) of 35 Root Cause Analysis (RCA) serious incident reports (‘Never Events’) from a NHS Foundation Trust.

The goal was to see if and how Safety-II/resilient healthcare principles could contribute to the quality of investigation, e.g. understanding misalignments between demand & capacity, adaptive capacity, and WAI vs WAD etc, in conjunction to systemic issues and then rate the quality of recommendations.

Providing background, it's noted that:

? Evaluations of RCA reports in healthcare have concluded that “most investigation recommendations are limited in scope, not robust and are not likely to be effective, a fact borne out by the continued occurrence of Never Events” (p2)

? This study drew on the CARE model (Concepts for Applying Resilience Engineering). CARE has a perspective that adaptations and adjustments to work as done (WAD) comes about due to fundamental misalignments between demands in the work system and the capacities available to meet those demands. That is according to CARE, “outcomes emerge from the interplay of misalignments and adaptations” (p2)

? Misalignments can occur for a variety of reasons, e.g. low staffing or high patient numbers may require staff to process patients more quickly (i.e. adaptation) until conditions return to normal; or adaptations due to missing equipment, low staff knowledge, underspecified procedures or poorly designed devices

Results

Overall, most recommendations in the RCA reports were found to be of low to moderate effectiveness for preventing issues in the future, with just one judged to be comprehensive. Nine main misalignment themes & 149 specific misalignments were identified in the RCAs.

Authors state: “on measures of analytic effectiveness and resilience there were no reports that achieved a higher score than 56%, and there were many problems identified in the reports” (p6).

On the latter, it was observed that despite many reports identifying misalignment factors that might lead to future incidents, in many cases there were no proposed actions to mitigate or remove the risks (44% of 144 misalignments had no action assigned).

The types of actions and their effectiveness are shown below:

Some examples of misalignments included things like:

? Policies: inappropriate policies not appropriate to the situations

? Staff issues: busy/overcrowded/tired staff, staff being called away or substituted during procedures, undertrained staff or given inappropriate level of responsibility

? Documentations: incomplete or counter-intuitive documentation, in many RCA reports it was difficult for the investigators to understand the event because documentation was incomplete

? Checklists/checking: checklists for preventing never events weren't fit for purpose, failure to use checklists or checking procedures not followed or not adequate

? Storage and labelling: inappropriate drugs and equipment or confusing labelling of drugs and equipment

? IT: lack of access to IT leading to inability to verify info and reliance on verbal info, lack of interoperability between systems

? Change implementation: Proposed actions/changes not implemented, lack of change control like lack of access to electronic patient records

Moreover, key main weaknesses in the RCA reports were:

1. Not understanding or describing WAD, including barriers for staff and how they solve problems
2. Not addressing misalignments that indicate weaknesses in systems, e.g. incomplete documents, staff shortages
3. Not considering how the identified problems could affect other areas or risks, how they manage the risks, or how to apply solutions elsewhere
4. Not addressing staff wellbeing
5. Lack of holistic and systematic learning after incidents, where instead new items are added to checklists etc
6. Policy problems left unaddressed, where few attempts made to understand how/why people follow (or not) policies.

Nevertheless, RCA reports did contain some descriptions of WAD and some reports also considered how issues and challenges were normally solved. Even with a “wealth of information about misalignments between demand and capacity” available from the RCA reports, these misalignments were often not resolved.

Interestingly, the authors note that current regulatory systems and techniques including SMART goals “discourage incident investigators from considering solutions that are outside their control, so they will seldom suggest actions that they do not have the power to implement” (p6). This lack of requisite imagination is said to result in more localised firefighting, rather than aspirational actions or upstream factors.

Importantly, the authors highlight that both S-I & S-II are complimentary and necessary for improving quality of care – rather than one or the other.

But, they say their findings highlight how using a S-II lens can improve opportunities for creating safer systems that are “lost in many RCA reports”, and improve investigations by directing investigators’ attention to better understanding healthcare work, WAI/WAD, misalignments and the need for adaptations to manage problems.

Moreover, investigation processes should be “used to understand systems and their weaknesses [34, 35], not only to understand the event that occurred” and thus, there appears to be many lost opportunities following never events for system learning and improvement; with instead a myopic focus on the exact event itself.

On the latter, by better understanding healthcare work efforts can be made to better manage/avoid adaptation.

Link in comments.

Authors: Anderson, J. E., & Watt, A. J. (2020).?International Journal for Quality in Health Care,?32(3), 196-203.