Investigative Techniques

Investigative Technique

Based on the initial discovery of an event, it is difficult to determine the causes and, later, the appropriate preventive action. Therefore, a thorough event investigation needs to be conducted to discover the antecedents and contributing factors.

The general goal of the event investigation is to gain as much information as possible to fully understand the event. In a cognitive interview, success depends on two factors:

memory
communication

An investigation has multiple feedback loops where information gained in one activity feeds back to a previous activity. The new information can result in changes to previous steps/understanding or it can add new pieces of information to previous steps.

The boundaries of the investigation include everything within the control and responsibility of the transfusion service or blood center. An example of an event outside the boundary would be a supplier that did not deliver reagents in a timely manner. The event needs to be recorded and monitored, but determining causes is not possible.

The investigator should have a general understanding of the organizational setting and the work processes of the transfusion service or blood center. Remember, however, that expertise and familiarity does not ensure an effective investigation because familiarity may blind someone to certain situations or tempt that person to make quick assumptions.

First, avoid unnecessary repetition by finding out what has already been investigated. The discoverer of the event has already described a short version of the event on the Event Discovery Report. Contact the discoverer to find out what has already been done.

General Investigation Tips

Conduct the investigation as soon as possible while the event is still fresh in people’s minds
Display a positive attitude toward the event investigation
Do not rush to conclusions or make assumptions
Review the applicable Standard Operating Procedures
Consider an investigation as a research project to gain understanding about the root causes
Use knowledge of previous events to get additional clues to causal factors
Ask a others to review the root cause analysis investigations for completeness and accuracy

The Five "W" Approach – Who? What? When? Where? Why?

Use the following five questions to fully explore the event:

Who was involved?

Knowing which people contributed to or were associated with the event creates the opportunity to gather more information. Getting the names of people should only be used for gaining more information—not for assigning blame to individuals. Knowing the types of people (RNs, MTs or level of experience) is also valuable when looking for trends in event data.

What happened?

This is the core of the investigation. All details that are relevant should be gathered, such as details that provide links to other information and/or that indicate necessary corrective action and/or that provide contributing facts.

When did it happen?

The time at which something happened can reveal important elements in the evolution of the event. For example, behavioral attitudes might be revealed by knowing the day of the week or if it was a holiday.

Where did it happen?

The place of the actual event often reveals important facts, such as physical layout or working conditions.

Why did it happen?

Asking "why" should reveal new information on a level closer to the root causes. Asking "why" repeatedly often reveals new information that would other not be uncovered. A general rule is to ask why five times of any consequent or antecedent event to increase eat likelihood of identifying root causes, especially the latent conditions that contribute to events.

Interviewing

A great advantage of the interview is the face-to-face setting. In addition to soliciting needed information, the interviewer can draw conclusions based on the way a person responds to a question.

Structured or Unstructured Interviews

A structured interview has a predefined set of questions and order, such as in a questionnaire. The focus of a structured interview is consistency in format and structure, which should be determined in advance. In fact, setting up a structured interview indicates that the exact information to be gathered has already been determined.

In contrast, an unstructured interview has no predetermined set of questions, which allows the interview to evolve based on the perception and direction of the interviewer and the memory of the interviewee. This type of interview is particularly useful when it is not yet completely clear what kind of information the interviewee can provide.

For both types of interviews, the following general guidelines should be applied:

Preparation

Prepare thoroughly for the interview. Establishing objectives and preparing questions or a list of topics to discuss keeps the interview on track and prevents the investigator from omitting key topics. Gather some background information about the interviewee. Understanding who you are talking to provides insight into the type of answers given or potential difficulties you might encounter. It also helps you know how much time to allocate for the interview. Always make an appointment and work with supervisors as needed.

Introduction

In addition to introducing yourself, make sure the interviewee knows the objective of the interview. Tell the person you only want information about what happened and are not looking for someone to blame. Also, let the interviewee know how long the interview will take.

Narration

Following the introductions, allow the interviewee to relate the event from the beginning in his or her own terms. If the person does not know where to start, use the following lead: "What first called your attention to the event?" This allows the interviewer to focus on his or her memory of the event as opposed to being directed to a specific element. It will also let the interviewee know that you are interested in his or her version. The goal of the narration is to develop a strategy for the remainder of the interview, not to collect specific details. Therefore, it is not important to record every detail at this point. All detailed answers will be gathered in the question portion of the interview.

Questions

After the interviewee finishes his or her story, you can follow up with questions that will clarify and amplify the event. (See "Interviewing Do's and Don’ts" below.)

Review

To check the accuracy of your notes and to provide an additional opportunity for the respondent to search through memory to uncover new information, review the interview by repeating key information gathered.

Interviewing Do's and Don’ts

Be mood-supporting, non-intimidating, non-judgmental. An interview is fact finding, not fault finding. You are searching for the root causes of the event. Be neutral but interested.
At the beginning, encourage interviewee to recreate the context of the event. This involves asking her/him to think about the context—where located, surroundings, feelings, and thoughts—and allowing five to 10 seconds for her/him to do so.
Use a private location. The presence of colleagues can decrease the accuracy of the answers.
Limit distractions.
Thorough note taking costs time during the interview and important information can be missed. If handwritten notes are used, rewrite the notes as soon as possible after the interview while details and abbreviations are still fresh on your mind.
Do not use terms the interviewee does not know. Consider differences in education or experience.
Organize questions so that each aspect of the event is examined at one time.
Avoid leading questions. Questions that suggest an expected answer will press the interviewee in that direction. Notice the difference between a leading question and an open-ended question:

"Was the blood type A positive?" vs. "What was the blood type?"

Asking open-ended questions increase the chances of getting additional details and greater accuracy from that person’s perspective. Closed questions can be helpful for directing attention toward relevant facts after basic information has been gathered.
Avoid negative wording: "Do you remember X?" rather than, "You don’t remember X, do you?"
Allow for pauses. Rapid fire questions interrupt memory processes and cause interviewee to limit responses because they feel rushed.
Ask several simple questions instead of one complex question. A complex question consists of several unrelated questions combined into a single question (e.g., "What time was it and what was he wearing and did you see his face?").
Do not interrupt the interviewee.

Collective Interviewing and the Building of a Causal Tree

One way to prevent bias from a single investigator is to build the causal tree collectively by including all the people involved in the event or those having a close relationship with the event.

First of all, the recommended size of the group is no more than eight people. In a group setting, the investigator should first explain the purpose of the meeting and briefly describe the event. If members of the group are unfamiliar with the causal analysis process, this should be explained briefly.

The investigator will build the tree on a surface that can be viewed by all participants so that everyone can contribute and comment during the process.

Collective tree building follows the same process as outlined in the section on building a causal tree. First, there must be agreement on the top event. If the event is a near miss, starting with the recovery side of the causal tree begins the discussion on a positive note. Next, the failure side of the tree can be constructed by asking the question "why" and "and why" as many times as necessary. A group working together can gather a significant amount of information and build the causal tree in a fairly short time.

Although collective tree building can be very effective, it has some potential drawbacks. Gathering all the appropriate people together at the same time can be challenging. Also, due to the presence of colleagues in the room, biased answers may be given and people may be reluctant to admit errors or failures in their departments.

An alternative to assembling a group is to create an electronic draft of the causal tree that can be emailed to all participants in the investigation. Each can add her/his own perceptions and email it back to the investigator. Several iterations may