# Feasibility and acceptability of hazard prediction training for potential hazard prevention at Champasak Provincial Hospital, Lao PDR: a case study

## TL;DR

This study explores how hazard prediction training can help improve patient safety in a hospital with a culture that discourages discussing past mistakes.

## Contribution

The study introduces hazard prediction training as a novel approach to initiate patient safety in a punitive healthcare environment.

## Key findings

- Participants identified 55% of the intended hazards using hospital scene photos.
- Environmental and software factors were most commonly recognized, while patient factors were overlooked.
- HPTs led to concrete actions like organizing medicines and posting posters in wards.

## Abstract

Learning from failure is recommended for the prevention of incidence recurrence. However, a punitive climate hindered organizational learning from past failures, especially in the early stages of patient safety initiatives. Little is known about how to initiate patient safety in a punitive climate. Hazard prediction training (HPT), which focuses on healthcare professionals’ discussion of potential hazards without referencing past failures, was introduced at Champasak Provincial Hospital in Lao PDR. This study examined the feasibility and acceptability of HPTs as a potential patient safety approach under a punitive climate.

In August 2019, 29 nurses were trained in HPT using four rounds of group discussions based on seven hospital scene photos to identify potential hazards, underlying factors, and priority countermeasures. A qualitative content analysis was performed on the discussion results. We analyzed differences between the 20 intended hazards and the participants’ responses, awareness of risk factors based on the P-mSHELL model, and adherence to the discussion methods.

The participants identified 55% (11/20) of the intended hazards. Five unintended hazards were identified mainly in complex scenes with many objects. Environmental factors were recognized most, followed by software factors, such as the absence of rules. Although patients’ families were identified as liveware factors, patient factors were overlooked. Discussions led to the identification of hazards and factors when the main subject in the photo was evident. The proposed countermeasures tended to be broad enough to cover all identified factors, and thus were abstract. Nevertheless, the result led to concrete actions such as tidying the warehouse, organizing the medicines and posting posters in the wards.

HPTs facilitated discussions on potential hazards in a setting where discussing past failures was culturally discouraged. Findings suggest HPTs are a feasible and acceptable entry point for patient safety initiatives in similar contexts. While not addressing past failures, HPTs may raise awareness of risks and shared responsibility. Frequent recognition of environmental factors suggests synergy with 5S (sorting, setting in order, shining, standardizing, and sustaining) activities. Given the single-site design and limited sample size, further research is needed to assess the broader applicability and impact of HPTs in diverse healthcare environments.