Factors Associated With Parent-Perceived Miscommunication in the Pediatric Intensive Care Unit

Clinicians who work in pediatric intensive care units (PICUs) provide immediate health care interventions to seriously ill children in a very stressful environment. Good communication between parents and health care professionals in the PICU is essential, but suboptimal communication is well documented.^1–6  According to the results of several qualitative research studies, suboptimal communication between clinicians and parents is sometimes insensitive, dehumanizing, vague, incomplete, or even perceived as dishonest.^1,2,5–7 

The trauma of seeing their child in a vulnerable state can be overwhelming for parents. Moreover, parental stress in the PICU may be linked to parent-clinician communication and the need for constant updates on the risks and benefits of any given treatment.^8,9  Many parents of hospitalized children experience severe anxiety (51%), major depression (24%), and traumatic stress during and after a PICU stay.^10–12  Parental stress in the PICU has also been linked to posttraumatic stress symptoms¹³  and can adversely affect memory retrieval.¹⁴  Parents, as experts on their child, can offer information about their child’s baseline status.³  Complex and urgent information exchanges between parents and health care professionals could lead to erroneous conclusions being drawn, with adverse consequences for the child.

In a study by Khan et al,⁴  miscommunication was reported by 15% of parents with hospitalized children and was significantly more common among parents of children with longer stays. Miscommunication between parents and clinicians in the PICU could be especially harmful given the stress of the environment, aggressive treatment options, and high potential for comorbidities in critically ill children.¹¹  Miscommunication within critical care environments can carry long-term health consequences and can add to the psychological distress that parents experience.^10,12,13 

The science of communication, particularly focusing on miscommunication, is understudied in the PICU. This lack of research has significant implications for patient safety and other patient- and family-related outcomes. Therefore, the aims of this study were to quantitatively describe parents’ perceptions of the frequency and types of parent-clinician miscommunication (eg, consistency, complexity, and accuracy) about their seriously ill children in the PICU and to examine the associations between parents’ perceptions of miscommunication in the PICU and parent demographic factors (eg, age, race, and ethnicity), parent psychosocial factors (eg, perceived level of stress, trust, and decision-making style), and child characteristics (eg, acuity, diagnosis, and length of stay [LOS]). Associations between perceptions of miscommunication and parent psychosocial factors were the primary outcomes of interest in this study.

We conducted a cross-sectional survey of parents of hospitalized children in the PICU. A parent was defined as an adult who was the primary caregiver for the ill child. Miscommunication was defined as the failure to communicate clearly as perceived by relevant stakeholders in the PICU, such as parents and health care professionals. The study was deemed exempt following review by the institutional review boards of the Children’s Hospital of Philadelphia (protocol 21-018963) and the University of Pennsylvania (protocol 849614). This study was conducted in accordance with the ethical standards of the Declaration of Helsinki.

Participants were recruited from a quaternary pediatric hospital and level I trauma center with a 60-bed PICU. A query of a local data set of the virtual PICU system at the designated hospital identified patients who were admitted to the PICU between January 1, 2018, and February 29, 2020, and had a PICU stay of longer than 24 hours. This time frame ensured that all potential parent participants had time to interact with health care professionals. Potential participants were excluded if they did not speak English, if the patient’s medical record did not include a parent contact or a valid phone number, or if the patient was deceased. We excluded these groups because they would likely have unique communication needs different from those of the general PICU population.

The local virtual PICU system data set provided a list of 7194 potential participants. Every ninth participant was selected on the basis of systematic random sampling, yielding 549 eligible participants. Of this total, 189 could not be reached, 95 chose not to enroll in the study, and 215 provided consent to participate. The research study was explained via telephone, and verbal HIPAA (Health Insurance Portability and Accountability Act) authorization was obtained. The institutional review board–approved consent form was sent with the survey via email, and the survey could be completed only if participants completed the consent form. Parents were instructed to consider their time in the PICU when responding. Only 1 parent in each family was eligible to participate.

For this study, an online questionnaire focusing on parent-perceived miscommunication in the PICU was developed and validated using published measures.^9,15–19  We conducted a pilot test of the survey with 10 parents and subsequently made minor wording changes for clarification.

Parent demographic information included age, race, ethnicity, sex, gender identity, relationship to the child, relationship status, household income, education, and insurance type. Child demographic information included age at admission, PICU LOS, number of diagnoses, acuity scores (Pediatric Index of Mortality 2 and Pediatric Risk of Mortality III), reason for admission, and number of previous PICU admissions.

Parent psychosocial factors included stress, trust in physicians, views of physician interpersonal communication skills, family functioning, and decisionmaking preferences. We measured these factors with the following validated instruments: the Parental Stressor Scale: PICU (α = 0.92),¹⁵  the Trust in Physician Scale (α = 0.89),¹⁶  the Communication Assessment Tool (α = 0.98),¹⁷  and 3 subscales of the Family Assessment Device (α = 0.92).¹⁸  All instruments used a 4- or 5-point Likert scale ranging from “strongly disagree” to “strongly agree” and scoring followed the authors’ guidelines. We used the Control Preferences Scale,¹⁹  a single item with 5 response options, to ask parents the degree of control they wanted over decision-making.

We measured parents’ perceptions of miscommunication in the PICU with a 6-item miscommunication scale with responses ranked on a 5-point Likert scale ranging from 1 (“strongly disagree”) to 5 (“strongly agree”). Higher summary scores indicated a greater degree of perceived miscommunication. The 6 items were adapted from the work of Khan et al⁴  and an existing study.⁸  Content and construct validity have been reported to have good internal consistency reliability (α = 0.89).⁹  We also included 9 items on the survey questionnaire to ascertain the type and frequency of miscommunication events (eg, conflicting information, delay in getting key information, incorrect information, etc), adapted from Khan et al,⁴  and several items to measure perceived stress for parents, family members, and the child.

For this study, a small correlation of r = 0.20 could be detected with 80% power at a 2-tailed α of .01 to detect bivariate relationships with the outcome measure of perceived miscommunication. Two hundred participants were required to power this study. Because of the potential for missing data associated with survey research and using a conservative 20% attrition rate, the target recruitment was 250 participants. The multiple linear regression models in the study would therefore have 80% power to detect an f² value of 0.066 to be significant at the α = .01 level.

Data were analyzed with SPSS version 28 (IBM). Descriptive statistics included measures of central tendency for all numeric variables and frequencies and percentages for categorical variables. Missingness, skewness, and kurtosis were assessed for each numeric variable, and transformations were made if indicated. We conducted mean imputation when less than 20% of items were missing. Parametric and nonparametric tests were used to evaluate the relationships between the independent variables and parent-perceived miscommunication. Before regression, the potential for multicollinearity was assessed on the basis of the variance inflation factor. Tests for significance were evaluated at the α = .01 level to reduce the risk of a type I error. Residual analyses were performed on the multiple linear regression models.

A total of 215 participants (59.7% of the 360 eligible participants we were able to reach) consented to participate in the survey. Most participants were female (83.4%), White (71.4%), and not Hispanic (87.9%). The median age was 39 years (mean [SD], 40.2 [8.75] years). Nearly 20% of participants were Black or African American and most participants (70.0%) were married (Table 1). Hospitalized children ranged in age from 15 days to 29 years; their median age was 3.83 years (mean [SD], 6.1 [6.02] years). Respiratory illness was the reason for 38.6% of admissions; surgical interventions were the reason for 33.8%. For 51.0% of patients, this was their first admission (Table 2). Mean (SD) LOS was 4.5 (6.2) days and mean (SD) Pediatric Risk of Mortality III score was 2.45 (4.05) (Table 2). Cronbach α values for all instruments used in the study ranged from 0.89 to 0.98 (Supplemental Table 1, available online only at ajcconline.org).

Reported levels of perceived miscommunication were low overall (miscommunication scale scores: mean [SD], 10.86 [5.06]; range, 6 to 28) (Supplemental Table 1). Of the 200 respondents, 33 (16.5%) agreed that miscommunication occurred during their child’s hospitalization. On the perceived miscommunication scale, 31 (15.5%) reported receiving inconsistent information, 23 (11.5%) reported having communication problems, and 10 (5%) reported receiving false or inaccurate information. When asked about specific types of miscommunication, 32 respondents (16.0%) reported receiving conflicting information and 20 (10.0%) reported a delay in receiving key information about their child’s illness (Supplemental Table 2, available online only). When asked how frequently miscommunication occurred in the PICU, 16 respondents (8.0%) reported a frequent lack of direct communication with their child’s primary health care professional and 11 (5.5%) also reported a frequent lack of consistency with the health care professionals caring for their child (Supplemental Table 3, available online only).

Bivariate correlations are shown in Supplemental Table 4 (available online only). Length of stay (r = 0.228, P < .001) and number of diagnoses (r = 0.144, P = .04) were positively correlated with parent-perceived miscommunication. Parent psychosocial factors that were significantly correlated with perceived miscommunication included higher perceived acuity of child’s condition (r = 0.168, P = .02), higher scores on Parental Stressor Scale: PICU (r = 0.465, P < .001), and lower scores for family functioning on the Family Assessment Device (r = 0.279, P < .001). Perceived miscommunication was inversely correlated with scores on the Communication Assessment Tool (r = −0.576, P < .001) and Trust in Physician Scale (r = −0.464, P < .001). We found no relationship between perceived miscommunication and parents’ decision-making preferences. Incidental findings include patient LOS being significantly correlated with both parental stress and Pediatric Risk of Mortality III score. Categorical demographic variables that were significantly associated with parent-perceived miscommunication, according to 1-way analysis of variance, were parental income (F = 2.804, P = .04) and parental education level (F = 2.977, P = .02).

Stepwise linear regression was conducted with variables significant at the level of P < .20 entered into the models. All resulting significant variables were then entered into an additional stepwise model. In this model, LOS and Trust in Physician Scale scores were no longer significantly associated with perceived miscommunication, but owing to the clinical significance of these variables, they were retained in a final model (Table 3). In the final model, the strongest predictors of parent-perceived miscommunication were scores on the Parental Stressor Scale: PICU (standardized ß = 0.286, P < .001) and the Communication Assessment Tool (standardized ß = −0.400, P < .001). The final model, which also included Trust in Physician score and LOS, accounted for 45% of the explained variance in parent-perceived miscommunication (R² = 0.448, F = 41.19, P < .001). Parental demographics were controlled for in the final model and had no meaningful impact on the significance of the variables or the explained variance of the model. We also conducted a sensitivity analysis to capture time since the PICU visit and examine potential recall bias and found no significant differences.

Understanding the factors associated with perceived miscommunication is an important step in understanding the scope of the problem and ways clinicians might help parents navigate what they often perceive as a chaotic and untimely situation for their child in an environment with high levels of stress. This study has 3 important findings. First, the degree of parent-perceived stress was associated with perceived miscommunication in the PICU. Second, the interpersonal communication skills of ICU clinicians are important to parents. Third, parental decision-making style was not a significant factor in miscommunication, although many parents wanted a shared approach.

Parental stress was significantly associated with parent-perceived miscommunication. Evidence suggests that many parents of patients in the PICU exhibit clinical levels of traumatic stress and those with pre-existing depression or anxiety report higher levels.^11–13  Stress can be physically and emotionally draining for parents and is associated with their perception of miscommunication. This link between miscommunication and parental stress offers a potential opportunity to mitigate or prevent parental stress through interventions that reduce parent-clinician miscommunication. This finding can help inform health care professionals on how they think about and converse with parents in the PICU and how parents’ degree of stress may impact their understanding of day-to-day events in the life of their hospitalized child. The mechanism by which stress influences perceived miscommunication in the PICU is unknown, but future research should examine whether other factors in the PICU, such as the sights and sounds of equipment, parental role alteration, the child’s appearance, or fear of death, may affect misunderstanding and miscommunication.^20,21  We also do not know whether an ethics professional or other type of consultative service professional was available for parents to discuss their concerns and how the availability of such a service might have mitigated their stress during the PICU stay. Parents may not know what resources are available to them when their child is admitted to the PICU. A simple list of all available resources could be provided or placed at the child’s bedside for their purview, providing them an opportunity to ask questions. It is also possible that miscommunication led to parental stress or that clinicians avoided parents who were showing signs of stress, thus increasing parents’ perceptions of miscommunication. Future research should explore this topic more fully.

A small but significant number of parents surveyed reported miscommunication, and 16.0% stated that this miscommunication came from conflicting information that was provided to them in the PICU. We also found that parental views of clinicians’ interpersonal communication skills were related to perceived miscommunication. Tarbi et al²²  noted that “high-quality communication can mitigate suffering during serious illness.” Given the prevalence of poor mental health outcomes for parents of patients in the PICU,^10–13  improving parental health must be a clinical and research priority. Perceived miscommunication can be a form of emotional or psychological suffering for parents of patients in the PICU and should be treated as a preventable adverse outcome. Qualitative studies also show that good communication increases trust^23,24  and poor communication decreases trust.^1,2,5,6,24  Although we found a bivariate association between trust and miscommunication in this study, this finding did not remain significant in the final regression model. Trust in the clinician, however, was associated with communication, and further work is needed to examine the relationship between trust and communication.

Shared decision-making in the PICU enables both health care professionals and parents to participate in the decision-making process, share their expertise, and communicate what they value.²⁵  Although perceived miscommunication and parental decision-making style were not correlated, understanding the degree of involvement that parents want to have in their child’s hospitalization is important. Inconsistency among clinicians in the PICU may contribute to miscommunication and lack of trust from parents. Decision-making can also be complicated when parents receive conflicting information. Additional study of the parent-clinician communication encounter, the value of clinician continuity, and how continuity may influence the decision-making process or parents’ decision-making style is needed.

Length of stay is a clinically important variable, although it was not statistically significant in the final model. Longer stays most likely provide parents with more communication encounters with health care professionals, thus either increasing parents’ probability of experiencing a miscommunication during the PICU stay or establishing consistent and trusting relationships. Other research has shown that children with more chronic health conditions have longer stays and higher mortality rates in the PICU than do children with fewer chronic conditions.²⁶  Parent’s ability to identify miscommunication events could also change with longer stays in the PICU. Length of stay was significantly correlated with both parental stress and Pediatric Risk of Mortality III scores, so this variable must be considered in future work.

This study requires consideration of several limitations. The study sample consisted of parents of children discharged from a PICU who were still alive at the time of the survey and were admitted to the PICU between January 1, 2018, and February 29, 2020. For parents, retrospectively examining a PICU admission increases the potential for recall bias but also allows time to reflect on the experience and recover from any acute stress related to the admission. We found that parents were very willing to participate and express their views. A small number of patients in the sample were older than 18 years and had parental caregivers during their PICU admission, a situation that can be common for individuals with complex health care needs who have not yet transitioned to adult health care professionals or are followed by pediatric specialists into adulthood.^27,28  This fact may limit the generalizability of our findings for PICUs that never treat patients older than 18 years. Future work should examine how older patients in the PICU might affect perceptions of parental miscommunication.

This study was conducted in a single PICU and would be strengthened by additional studies in different geographic locations. A strength of the study was that 28.5% of the parent respondents were not White, but more racially diverse samples are necessary to evaluate potential disparities related to perceptions of miscommunication in the PICU. We excluded parents whose children died and also those who did not speak English; however, in future research parents in these subgroups might provide unique insights into perceptions of miscommunication and experiences within the PICU. Translating and psychometrically testing the miscommunication scale are needed to assess its reliability and validity in different parent subgroups. Finally, although our focus was on parents, who provided valuable data, knowing what clinicians believe they can do to make parents’ experiences better is also important. Understanding clinicians’ views of what constitutes good communication in the PICU would add to our findings.

Given the complexity of communication in the PICU, understanding what is said, how it is received, and how it affects patients and their families is critical to advance pediatric communication research.²²  The stress of the PICU environment is real and can be long-lasting. Decreasing perceived miscommunication may mitigate this stress and help parents and clinicians communicate in a way that supports the hospitalized child, builds trusting relationships, and helps them make decisions that reflect mutual preferences and goals. More dialogue would be helpful to explore the importance of clinician consistency in the PICU, resources that support parents during their child’s illness, and how shifting parental preferences and goals during interactions with health care professionals may influence perceptions of miscommunication.