Madison Patient Safety Collaborative

Making Madison the safest place to receive healthcare

Patient Falls Initiative - Aggregate Root Cause Analysis

Building on successes from our initial fall reduction project, a proposal for an Aggregate Root Cause Analysis for falls was created for and subsequently accepted by the MPSC on February 9, 2004, and a working group was formed.

Project Goals/Objectives and Timeframes

The primary goal of this project was to:

improve the safety of hospitalized patients in Madison, WI, by decreasing the number of patient falls and injuries from falls

Secondary goals included:

deepening participants’ knowledge and skill with the National Center for Patient Safety’s technique for Root Cause Analysis and Aggregate Root Cause Analysis
using a single Aggregate Root Cause Analysis to include events from four different hospitals
evaluating the practicality, benefits and barriers to doing so

The Aggregate Root Cause Analysis was completed May 1, 2004 and four root cause statements were identified.

Methods

Investigational Review Board (IRB) approval was obtained from each hospital. Data was collected on 15 falls on the medical services and 15 falls on the surgical services on scannable forms. Results were entered into a database.

At the first Aggregate Root Cause Analysis (ARCA) meeting, following a brief overview of the proposed process, the literature on falls involving hospitalized patients was reviewed. A falls flowchart was developed and discussed in relation to the available data that could be further analyzed during the discussion. Participants agreed to focus initial efforts on the assessment process because 46% of the patients who fell had not been assessed as being at high risk for a fall. An initial cause-effect diagram was developed.

Following review and refinement of the cause-effect diagram at the second meeting, root causes were developed and finally, utilizing the five rules of causation, root cause statements were written.

At the final meeting, root cause statements were reviewed, interventions and outcome measures developed, and options for elimination, control, or acceptance of the root cause were discussed. The four root cause analysis statements are:

The risk assessment may be inaccurate due to a tool with high specificity, but limited in sensitivity thereby increasing the potential for falling not being detected.
The triggers for reassessing patient with change in condition (better or worse) are not clearly identified/understood, thereby increasing the risk of falling.
Due to barriers of communication of patient information regarding risk factors for falling, the assessment may be inaccurate leading to an unrecognized risk for falls.
The scope of the fall risk assessment training program increased the likelihood that all risk factors for falling will not be considered in the assessment, leading to an inaccurate assessment and increased risk for falling.

Implementation

To address the root causes the following actions were implemented
Fall risk assessment tools were redesigned to include an override of the fall assessment based on the nurse's judgment.
Examples of a "patient change in condition" were added to the assessment tools to serve as concrete situations as to when a reassessment should be done.
Some units implemented reassessments every 8 hours.
A survey was distributed at each hospital to capture data regarding information flow and communication barriers.
The team is currently working on planning a falls conference to disseminate fall reduction information.

Conclusions and Outcomes

Conclusions regarding the effect of this effort on the reduction of falls and injury from falls are deferred until interventions have been implemented. The above mentioned final team meeting occurred on April 22, 2004.

However, preliminary comments on the secondary goals are possible. Although data were not collected, participants’ knowledge of the processes of RCA and ARCA was deepened. The presence of four different hospitals using four different event reporting forms, four different assessment tools, and four different sets of interventions for those at high and normal risk for falls slowed the process due to the multiple explanations required. However, agreement could be reached for a common data collection tool, a focus for analysis and interventions and the interventions themselves. In addition, the sharing of ideas due to the above differences demonstrated the benefits and excitement associated with collaboration.

The group is currently working on preparing the collaborative process, implementation and outcomes of this project for publication.