In 2010, a team of educators and clinicians at Scripps Mercy Hospital in San Diego, California, USA, undertook a project to enhance communication with families of patients in the intensive care units (ICUs), based on the premise that decision making in the ICU requires a shared model, and that effective shared decision making requires clinicians who are trained in meeting management, conflict management and communication skills. It was proposed that clinicians be trained to develop these skills in multiprofessional teams to encourage the development of a common voice for responding to patient and family concerns.
Baseline family satisfaction with communication, and staff confidence in providing that communication, were measured from September 2010 to March 2011. Over a period of four months (May 23 through September 19, 2011), 14 training sessions were held for a total of 98 clinicians; each session held between six and 10 individuals from multiple disciplines, including intensive care physicians, medical residents, nurses, social workers, case managers, pharmacists, respiratory therapists, and chaplains.
Training consisted of four steps, including:
• Reading articles about the Buckman SPIKES (setup, perception, invitation, knowledge, emotions, strategy or subsequent) protocol(1) and about patient and family definitions of high-quality care in the ICU, as well as a list of unit-specific verbatim comments from families of patients from our own ICU, describing both positive and negative communication experiences. We felt that the quotes would be more powerful than a literature review of family needs because they came from internal cases. Although there were many positive comments, reading the testimonies from actual patients made it clear that there were opportunities for improvement, stimulating an impetus for change. The reading assignment was done as homework before the training day. The majority of the participants had done the reading assignment before the training.
• A brief didactic presentation of the Buckman SPIKES protocol.(1) This exercise took about 30 minutes.
• A case simulation exercise (role play using a standardized patient/family set kindly provided by Robert Arnold, MD [R. Arnold, personal communication, September 2009]), in which each of the team members played one of the following roles: family member one, family member two, resident, intern, bedside nurse, charge nurse, chaplain, respiratory therapist, and intensivist. This exercise took about 30 minutes.
• A self-assessment debriefing of the case simulation by all trainees with the trainers serving as facilitators. This took about 30 minutes.
Post-intervention family satisfaction and staff confidence were measured from October 2011 to March 2012. Staff confidence was measured using a validated survey adapted from one provided by Walter Baile, MD. Family satisfaction with communication was measured using individual questions from the Family Satisfaction in the ICU survey (FS-ICU 24).(2)
Staff confidence was measured in paired questionnaires from 46 clinicians in whom pairing could be identified. For each of 21 separate measures of staff confidence, the improvement in confidence from pre- to post-intervention was highly significant (p<.001). There was no significant difference in outcome when unmatched questionnaires were compared with matched questionnaires.
Quantitative Analysis: Family Satisfaction
A family satisfaction survey (FS-ICU 24) was mailed to the closest family member listed in the patient’s demographic information. The form was sent four to six weeks after the patient’s discharge or death. Return rates were 12.3% before and 15.3% after the intervention. Using standard statistical methods, results were compared on seven survey questions related to communication. Three of those questions showed significant improvement after the intervention, while the remainder showed no significant change in satisfaction.
Qualitative Analysis: Family Satisfaction
One of the most interesting findings in the study was actually removed from our published results(3) because of space limitations, and we are pleased to have the opportunity to present it here. A qualitative analysis of write-in comments on the FS-ICU 24 survey was performed independent of the quantitative scoring using general qualitative methods. The survey was abstracted for the main topic measured within each question. These topics were placed in a table. Each write-in comment was then coded as a positive or constructive statement and matched to the closest topic. If the comment did not match any topic, it was sorted to “other.” If the comment was written about an experience outside of the ICU, it was not counted. If a respondent commented more than once about a given topic, the item was only counted once. A comment might have been coded into more than one construct if it referred to multiple constructs, such as the skill and competence of nurses plus concern and caring. If the comment was about caring and it could not be discerned whether the caring was focused on the patient or on the family member, it was coded to “concern and caring to your loved one.” Comments about excellent care were coded to “skill.” The counts of positive and constructive comments were then tallied in three ways:
• All comments (including those related to support during death and those coded to “other”)
• Part 1 satisfaction with care questions only
• Part 2 satisfaction with decision-making questions only
A chi-square analysis was performed to determine whether there was a significant difference between the pre- and post-intervention results in any of the three groups. Due to the small number of comments received for each question, analysis at the individual question level was not appropriate and not attempted.
No significant difference was found in the volume of positive to constructive comments in either the entire data set or part 1 satisfaction with care questions. This mirrored the findings from the quantitative analysis. However, just as was found with the traditional scoring of the FS-ICU 24 survey, a significant difference was found in the proportion of comments related to part 2 satisfaction with decision making (x2 = 12.01, P<.001). Often managers read the write-in comments on surveys individually but do not take the time to analyze them for theme. We have found that this can be done without much effort.(4) The qualitative analysis lends further validation to the scores and can be very helpful in helping staff to understand the results because of the specific examples.
Also, and importantly, the “other” section accumulated comments reflecting three communication topics of importance to family members that were not evaluated on the FS-ICU 24 survey. These included timeliness of information received, appropriateness of communication in the patient care area, and comportment (professional versus rude communication).
The results of the qualitative analysis further supported the quantitative findings that our team training improved family satisfaction with decision making. Most of the elements of the section on decision making evaluated items related to method and quantity of communication, which was the focus of the intervention. Because the survey section on satisfaction with care focused on items such as skill, caring, ICU atmosphere, or waiting room atmosphere – and this was not the focus of the intervention – it is easy to explain why the comments did not significantly change within this portion of the survey. It is known that how we communicate is just as important as what we communicate.(5) Given this, the new items of importance identified through qualitative analysis regarding timely, appropriate and professional communication could be used to further modify and strengthen the FS-ICU 24 survey.
Translation into Practice
Following the study, we attempted to translate the findings into practice by standardizing the training throughout the ICUs in our four other hospitals. We were able to replicate the project successfully in one ICU, which was staffed by these same residents and nursing leadership. However, due to cost containment, efforts were thwarted when attempting to implement the findings in the other three hospitals within the healthcare system, which do not have residency programs. Internal funding was not available to compensate nurses for attending the two-hour training program. In today’s healthcare environment, these budgetary restrictions are universal. Future research should focus on meeting the learning needs of staff in an economically challenged healthcare environment. New residents continue to receive the training, and the ICU managers at the hospitals where the transition into practice was successful are invited to send nurses to the training program to keep up with turnover. Nurses do attend, but there is room for improvement in their participation rate.
We conclude that a 90-minute interactive training session based on the Buckman (SPIKES) protocol, coupled with prereading, a list of actual ICU-specific family comments, and a case simulation involving all trainees, improved family satisfaction in key areas of ICU decision making that relate to communication. The attendance rate (52.7% of invitees) suggests that the training was positively received. A posttest survey demonstrated a significant improvement in clinician confidence with communication. A larger multicenter trial is needed to further validate results and broaden the ability to generalize.
1. Baile WF, Buckman R, Lenzi R, Glober G, Beale EA, Kudelka AP. SPIKES-A six-step protocol for delivering bad news: application to the patient with cancer. Oncologist. 2000;5(4):302-311.
2. Wall RJ, Engelberg RA, Downey L, Heyland DK, Curtis JR. Refinement, scoring, and validation of the Family Satisfaction in the Intensive Care Unit (FS-ICU) survey. Crit Care Med. 2007;35(1):271-279.
3. Shaw DJ, Davidson JE, Smilde RI, Sondoozi T, Agan D. Multidisciplinary team training to enhance family communication in the ICU. Crit Care Med. 2014;42(2):265-271.
4. Davidson J, Lamontagne G, Burnell L, et al. Analysis of staff safety concerns. J Nurs Care Qual. 2013;28(2):147-152.
5. Siegel MD, Hayes E, Vanderwerker LC, Loseth DB, Prigerson HG. Psychiatric illness in the next of kin of patients who die in the intensive care unit. Crit Care Med. 2008;36(6): 1722-1728.