RESEARCH IN CONTEXT
- Adult ICU research surrounding prognostic communication highlights differences in how surrogate decision-makers interpret language used during prognostic conversations by providers.
- Neonatal ICU research demonstrates differences in provider and parental perception about a child’s prognosis.
- There is a need to explore prognostic communication in the PICU. This study describes parent and PICU attending reports of prognostic discussions and parental interest in receiving more prognostic information.
Parents value clear communication from healthcare providers about their child’s medical condition(s) (1–3). Communication with parents should include information about prognosis when appropriate. Prognostic conversations are discussions during which providers share predictions about possible outcomes. Most knowledge about prognostic conversations between providers and surrogates of critically ill patients comes from literature about adult (4–7) and neonatal ICUs, with a focus on mortality (8,9). In the PICU, prognostic conversations may include information shared with families about PICU length of stay (LOS), as well as short-term morbidities, post-PICU morbidities, and mortality (10–12). Morbidities after an ICU stay are termed postintensive care syndrome (PICS) and encompass physical, cognitive, and psychologic outcomes that can affect the patient and their family (13). Recent reviews note that parents of PICU patients request more education and communication from providers about potential post-ICU outcomes (10–12). However, little is known about the percentage of PICU patients whose parents have prognostic conversations with providers or what content is addressed during such conversations.
The aim of this study was to evaluate the occurrence of prognostic conversations in the PICU based on parent and attending physician report. We focused on conversations pertaining to PICU LOS, PICU mortality, and PICS outcomes, specifically physical, cognitive, and psychologic morbidities for patients and psychologic morbidity for parents. We hypothesized that most parents would not report having prognostic conversations, and there would be poor agreement between parent and PICU attending report about the occurrence of prognostic conversations. Additionally, we explored the topics discussed during prognostic conversations, which healthcare providers discussed prognoses with parents, and whether parents wanted more prognostic information about their child and themselves.
We conducted the study from September 2019 to August 2020 at a 40-bed university-based PICU in Chicago, IL. Patients in the PICU are cared for by bedside nurses, advanced practice nurses (APNs), residents, hospitalists, fellows, attending physicians, subspecialty consultants, respiratory therapists, nutritionists, pharmacists, social workers, and chaplains. The study was reviewed by the hospital’s institutional review board (IRB) and approved with waiver of signed consent (IRB 2019-3075, approved July 25, 2019).
Study participants included PICU attendings and parents of patients less than 18 years old with multiple organ dysfunction syndrome (MODS) within 24 hours of PICU admission. These patients have longer hospital LOS and are at higher risk of mortality and morbidity (14–17), thus seem more likely to warrant prognostic conversations. We included biological parents, foster-parents, or legal guardians with decision-making authority. If both parents were present, one of the two parents completed the survey. We excluded non-English or Spanish-speaking parents or parents who already participated in this study. We also excluded parents if their nurse or attending physician felt it inappropriate to approach them or if the Department of Child and Family Services was involved. PICU attendings received surveys after their patient’s parent completed one. We surveyed PICU attendings because they are responsible for PICU patients’ care.
Study data were collected and managed using Research Electronic Data Capture (REDCap) electronic data capture tools hosted at Northwestern University Feinberg School of Medicine (18). REDCap is a secure, web-based application designed to support data capture for research studies. Parents and attending physicians completed surveys in English or Spanish via REDCap using an e-mail link or an iPad provided by the research team within 5–10 days of patients’ PICU admission. We chose this timeframe because: 1) the average PICU LOS at our institution is 5–6 days, 2) goals of care discussions occur on average by PICU admission day 2 (19), 3) it gave each family the same timeframe to have prognostic discussions, and 4) we hoped recall bias would be minimized. Waiting until PICU discharge would provide families with longer PICU stays, more opportunity to have prognostic conversations, and could increase recall bias.
Surveys were developed with input from three pediatric intensivists about PICS-related topics (11,13). We conducted cognitive interviews to assess face validity with a PICU chaplain, social worker, and four parents; surveys were revised accordingly (20). Parent surveys focused on six topics relating to their child’s prognosis: PICU LOS; mortality; and post-PICU physical, cognitive, and psychologic morbidity. Surveys also evaluated conversations about parents’ post-PICU psychologic morbidity (Supplement 1, http://links.lww.com/PCC/B789). For each topic, we asked parents if prognostic conversations occurred, what was discussed, with whom, and if they wanted more prognostic information. Only parents who reported having mortality-related conversations were asked if they wanted more information about mortality. Parent surveys also included demographic information.
Physician surveys included parallel questions related to each prognostic topic (Supplement 1, http://links.lww.com/PCC/B789). On service-PICU attendings were emailed the survey link within 24 hours of parent survey completion. Physicians completed a one-time survey regarding demographic information.
We collected patient demographics and PICU hospitalization data from the electronic medical record.
We used descriptive statistics to present parent and physician information about prognostic conversation occurrence for each topic as well as for any prognostic outcome discussed. We present this information as a percentage of the total responses with 95% CIs. Our sample size goal was 100 based on study time limitations, anticipated enrollment, and anticipated range of 95% CIs (CI) from 0.12 to 0.18 for varying proportions from 0.10 to 0.30.
An adapted Kappa statistic evaluated concordance between parent and physician reports about the occurrence of prognostic conversations for each topic and any prognostic outcome. An adapted Kappa was used to account for the dependence between provider survey responses, as PICU physicians completed surveys for multiple patients. We used a bootstrapping method to estimate the se for each Kappa statistic. For parents and physicians, we combined “No” and “I don’t know” responses about the occurrence of prognostic conversations related to each topic based on two assumptions: 1) if someone did not know a conversation occurred, it is less likely to have happened and 2) it is unlikely for someone to have knowledge about what was conveyed during a conversation if they cannot indicate with certainty that a conversation occurred. For physicians, we combined the “Yes” responses, “Yes, I was the one who spoke to the family…,” and “Yes, someone else spoke to the family…”.
Using descriptive statistics, we report demographics, clinical information, reported outcomes discussed, and parent responses about wanting more prognostic information.
We screened 812 patients, and 137 met inclusion criteria. Of 110 parents approached, 103 agreed (response rate: 94% of those approached, 75% of those meeting inclusion criteria), and 101 surveys were analyzed (Fig. 1). Parent and physician surveys were completed at 5.9 days (sd 1.5) and 7.85 days (sd 2.8) after PICU admission, respectively. Supplement 2 (http://links.lww.com/PCC/B790) summarizes patient and parent demographics. Most admission diagnoses included a respiratory problem (40%), sepsis (30%), or a nontraumatic neurologic condition (13%) and had two identified dysfunctional organs (57%). The average PICU LOS was 18.5 days (sd 18.6), and 10% of patients died. Most parents were female (74%) and had prior ICU experiences (60%). Only 39.6% of parents were White, and 34% Latinx. Twenty PICU attendings participated: 11 (55%) female, seven (35%) less than 5 years post-fellowship training, six (30%) were 5–10 years post-fellowship training, and seven (35%) were greater than 10 years post-fellowship training.
Parent and Physician Report of Prognostic Conversations
Table 1 shows parent and physician reports of prognostic conversation occurrence by topic. Eighty-eight percent of parents (95% CI, 81.5–94.3%) and 82% of physicians (95% CI, 74.7–89.6%) reported having prognostic conversations about any topic. Supplement 3 (http://links.lww.com/PCC/B791) includes proportions and corresponding 95% CIs of parents and attendings who responded “Yes” to each topic. Prognostic conversations most reported by parents and physicians were LOS and post-PICU physical morbidity. Prognostic conversations least reported by parents were mortality and post-PICU patient psychologic morbidity. Physicians least commonly reported having patient and parent psychologic post-PICU morbidity conversations. Over half of the attendings did not know whether psychologic morbidity conversations occurred.
TABLE 1. -
Parent and Medical Doctor Report of Prognostic Conversations by Topic
||Yes, n (%)
||No, n (%)
||I Do Not Know, n (%)
|Length of stay
|Patient post-PICU physical morbidities
|Patient post PICU cognitive morbidities
|Patient post-PICU psychologic morbiditiesc
|Parent post-PICU psychologic morbidities
MD = medical doctor.
aResponses came from the 20 attending physicians caring for patients during the study. Some physicians completed a survey about more than one patient.
bMissing one parent data point.
cMissing two parent data points.
Parent and Physician Concordance
Parent and physician concordance for reporting the occurrence of prognostic conversations about any topic was fair (Kappa = 0.22; 95 % CI, 0.004–0.46). For each outcome, parent and physician concordance was 0.04 (95% CI, –0.1 to 0.2) for LOS; 0.25 (95% CI, 0.04–0.47) for mortality; 0.21 (95% CI, 0.02–0.33) for physical morbidities; 0.25 (95% CI, 0.015–0.42) for cognitive morbidities; 0.17 (95% CI, –0.12 to 0.45) for patient psychologic morbidities; and 0.14 (95% CI, –0.01 to 0.31) for parent psychologic morbidities. Supplement 4 (http://links.lww.com/PCC/B792) provides further details regarding concordance about each topic.
Outcomes Discussed During Prognostic Conversations
Table 2 describes specific outcomes discussed during prognostic conversations for each topic, reported by parents. Within post-ICU physical morbidities, parents were asked to further clarify discussions regarding new medical technology and changes in everyday activities. For new medical technology, parents recalled discussions about respiratory technology (n = 11), feeding equipment (n = 17), and need for dialysis (n = 7). Regarding changes with everyday activities, parents recalled discussions about nutrition (n = 13), weakness (n = 10), walking (n = 5), balance (n = 4), seeing (n = 5), and hearing (n = 1). Supplement 5 (http://links.lww.com/PCC/B793) includes “Other” responses, when parents provided free text.
TABLE 2. -
Parent Report of Outcomes Discussed During Prognostic Conversations Based on Each Topic
||Parent Report, n (%)
|PICU length of stay (n = 68)
||Discussed, but unable to predict
|Patient post-PICU physical morbidities (n = 49)
||New home technology
|Changes in everyday activities
|Patient post-PICU cognitive morbidities (n = 23)
|Patient post-PICU psychologic morbidities (n = 13)
|Parent post-PICU psychologic morbidities (n = 28)
PTSD = posttraumatic stress disorder.
Medical Providers Involved in Prognostic Conversations
Parents reported that physicians/APNs most commonly had prognostic conversations about LOS, mortality, physical, and cognitive morbidities (Table 3). Most parents either did not recall with whom they discussed post-PICU patient psychologic outcomes or they selected “Other,” with the option to respond with free text. Discussions with parents about their own psychologic outcomes most commonly occurred with a social worker and/or Chaplain. Parents reported frequent involvement of bedside nurses in discussions pertaining to LOS, physical morbidity, and parent psychologic outcomes.
TABLE 3. -
Parent Report of Medical Providers With Whom Prognostic Discussions Occurred
||PICU Patient Length of Stayn = 68 (%)
||PICU Patient Mortalityn = 18 (%)
||Post-PICU Patient Physicaln = 49 (%)
||Post-PICU Patient Cognitiven = 23 (%)
||Post-PICU Patient Psychologicn = 13 (%)
||Post-PICU Parent Psychologicn = 28 (%)
|Physician/advanced practice nursea
|I do not know
a“Physician” includes resident, fellow, or PICU attending.
bTransplant team, liver team, surgical team, otolaryngologist, oncologist.
cGenetics, surgeon, oncologist.
Parental Request for More Prognostic Information
Table 4 shows the number of parents requesting more information about each topic. A larger percentage of parents requested additional information when a particular topic was not discussed or if they had no prior ICU experiences. Post-ICU physical morbidities and LOS were the topics about which most parents requested more information. Parent requested more information about their own post-ICU psychologic morbidities least commonly across all categories.
TABLE 4. -
Parents Requesting More Information
||All Parents (N = 101)n (%)
||Topic Not Discussedn (%)a
||Topic Discussedn (%)a
||No Prior PICU Admission (n = 51)n (%)
||No Prior PICU Admissions (N = 60)n (%)
|Length of stay
|Patient post-PICU physical morbidities
|Patient post-PICU cognitive morbidities
|Patient post-PICU psychologic morbidities
|Parent post-PICU psychologic morbidities
aThe total number of responses changes for each topic based on parent report of whether a topic was discussed.
bFor mortality n = 18 since we only asked parents who reported having discussed mortality if they would like more information about it.
This prospective study describes PICU prognostic conversation occurrence based on parent and physician report. Eighty-six percent of parents and 80% of physicians reported having any prognostic conversations. Parents and physicians most commonly reported conversations about PICU LOS and post-PICU physical morbidities; parents and physicians least commonly reported discussions about post-PICU patient and parent psychologic morbidity. Overall, parent and physician concordance was fair (Kappa = 0.22) for any prognostic conversation, with a range from slight (Kappa = 0.04, LOS) to fair (Kappa = 0.25, post-PICU cognitive morbidities). Although parents reported receiving most information from nurses and physicians, the majority recalled discussing post-PICU parent psychologic morbidities only with social workers and chaplains. Most parents who did not report having conversations about LOS and post-PICU physical outcomes wanted more information about these topics.
Families reported having prognostic discussions about PICU LOS and post-ICU physical outcomes most frequently. In the adult ICU, proactive communication about LOS improves family satisfaction and decreases LOS and readmission rates (21–23). Communication experts highlight the importance of discussing LOS to reduce burden related to posthospital family obligations (24,25). In regard to physical outcomes, 75% of PICU patients may experience lasting physical morbidities (11,26). Parent recall of respiratory and feeding technologies may be due to the frequency of respiratory-related admission diagnoses and prolonged intubation. Although LOS and physical morbidities were the prognostic topics most frequently recalled, parents still frequently requested more information about these issues, suggesting opportunities for improving communication.
AT THE BEDSIDE
- Parents generally want more prognostic information, even if they recall having had prognostic conversations.
- Parents and physicians do not agree on what was said during prognostic conversations.
- There are many opportunities to improve prognostic communication within the multidisciplinary PICU team and with families.
Although 65% of patients and 85% of parents experience psychologic morbidities post ICU stay (27), low parent report of psychologic conversations suggests another opportunity for improving provider communication. Studies show that both prehospital mental health conditions and in-unit events (loud alarms, invasive procedures, prolonged LOS, and mechanical ventilation) are associated with post-ICU psychologic morbidities (10,11,28). Currently, no universal guidelines support mental health screening in the PICU; however, screening could identify at-risk parents and prompt clinicians to address and provide support for post-PICU psychologic outcomes (12,29,30). Mental health screening and provider education have the potential to improve how we communicate with parents about possible post-ICU psychologic morbidities.
For most prognostic topics, we found either slight or fair concordance between parent and physician report. Although we cannot fully explain this finding, we propose that limited communication between healthcare providers may be an important factor. Parents reported discussing psychologic outcomes with chaplains and social workers, and attendings generally did not have awareness about those conversations. This knowledge gap about prognostic discussions suggests that improving communication between multidisciplinary care teams might increase prognostic conversation awareness. Some might argue that it is burdensome, unreasonable, or may cause information overload for an attending to know about all conversations that occur about their patients. However, we contend that prognostic conversations are vital to parents’ understanding of their child’s condition, and thus, attendings should have some awareness about prognostic conversations that occurred. Another notable finding in our study showed fair concordance (Kappa = 0.25) between parents (18) and physicians (30) regarding mortality. Potential reasons for this finding are that physicians’ language was unclear, so parents did not recall mortality discussions; or, families may consciously or unconsciously avoid recalling them to self-protect. Literature shows that many providers lack training in discussing mortality; perhaps relatedly, providers tend to use euphemisms rather than the word “death” and rarely allow families enough time to respond during difficult conversations (31–33). Our findings suggest the need for improved communication about prognostic conversations among the multidisciplinary team rotating in the PICU and more training on how to effectively lead difficult conversations.
We acknowledge some important limitations to this study. Because we relied on parent and physician recall, we cannot know if our data accurately reflect prognostic conversation occurrence. We chose this approach because it was not feasible to observe/document all prognostic conversations in real time and because a parent’s perception of having conversations may be as important as reality. Similarly, parent reports of who had prognostic conversations with them were potentially subject to recall bias and error, as parents may not have known the difference between PICU providers. Additionally, we did not determine whether the patients’ clinical conditions warranted having prognostic conversations, as there is no standard of care for this measure; however, since 31.5–75.5% of parents wanted more prognostic information, our study suggests that regardless of clinical status, talking to parents about potential morbidities may be important.
Other limitations are worth noting. Our study only included patients admitted to the PICU with MODS, impacting generalizability. By collecting data at days 5–10 of PICU admission, we likely underestimated the number of conversations that occurred later in the hospital stay (average LOS for the study population was 18.5 d). The survey asked parents to recall prognostic conversations about predefined topics; thus, we could have missed other important prognostic discussions. Also, parents and physicians did not complete the survey at the exact same time, potentially impacting concordance. Combining “no” and “I don’t know” responses based on assumptions may have misled our concordance analysis. We did not collect data on the four patients deemed not approachable by their RN/medical doctor, thus cannot evaluate whether this introduced selection bias. We did not survey all providers who potentially had prognostic discussions, instead relying on attending awareness. Finally, we did not perform statistical comparisons of parent/patient characteristics with prognostic conversations report due to sample size. Future works should enroll more parents and providers to allow for statistical comparisons, consider reports from parents and multidisciplinary providers at the time of discharge, and include more open-ended questions and/or qualitative interviews.
Most parents and physicians reported having prognostic conversations. Although parents most commonly recalled prognostic conversations about PICU LOS and post-PICU physical morbidities, approximately half wanted more information about these topics. Parents were least interested in information about their own psychologic morbidity. This study raises important questions about when clinicians should have prognostic conversations with parents of PICU patients, who should take responsibility for having such conversations, and how the multidisciplinary care team can best be made aware when such conversations occur. More research is needed to help answer these questions related to prognostic conversations.
We thank Dr. Erin Talati Paquette for her helpful review of earlier article revisions.
1. Meert KL, Eggly S, Pollack M, et al. Parents’ perspectives on physician-parent communication near the time of a child’s death in the pediatric intensive care unit. Pediatr Crit Care Med. 2008; 9:2–7
2. Williams CN, Eriksson C, Piantino J, et al. Long-term sequelae of pediatric neurocritical care: The parent perspective. J Pediatr Intensive Care. 2018; 7:173–181
3. Brand SR, Fasciano K, Mack JW. Communication preferences of pediatric cancer patients: Talking about prognosis and their future life. Support Care Cancer. 2017; 25:769–774
4. Quinn T, Moskowitz J, Khan MW, et al. What families need and physicians deliver: Contrasting communication preferences between surrogate decision-makers and physicians during outcome prognostication
in critically ill TBI patients. Neurocrit Care. 2017; 27:154–162
5. Schuster RA, Hong SY, Arnold RM, et al. Do physicians disclose uncertainty when discussing prognosis in grave critical illness? Narrat Inq Bioeth. 2012; 2:125–135
6. White DB, Engelberg RA, Wenrich MD, et al. The language of prognostication
in intensive care units. Med Decis Making. 2010; 30:76–83
7. Chiarchiaro J, Buddadhumaruk P, Arnold RM, et al. Quality of communication in the ICU and surrogate’s understanding of prognosis. Crit Care Med. 2015; 43:542–548
8. Meadow W, Cohen-Cutler S, Spelke B, et al. The prediction and cost of futility in the NICU. Acta Paediatr Int J Paediatr. 2012; 104:397–402
9. Singh J, Lantos J, Meadow W. End-of-life after birth: Death and dying in a neonatal intensive care unit. Pediatrics. 2004; 114:1620–1626
10. Atkins EK, John M, Colville G. Families’ experiences of life in the year after a child’s critical illness: Navigating the road to a “new normal”. J Pediatr Intensive Care. 2020; 9:188–195
11. Watson RS, Choong K, Colville G, et al. Life after critical illness in children-toward an understanding of pediatric post-intensive care syndrome. J Pediatr. 2018; 198:16–24
12. Hartman ME, Williams CN, Hall TA, et al. Post-intensive-care syndrome for the pediatric neurologist. Pediatr Neurol. 2020; 108:47–53
13. Manning JC, Pinto NP, Rennick JE, et al. Conceptualizing post intensive care syndrome in children-the PICS-p framework. Pediatr Crit Care Med. 2018; 19:298–300
14. Typpo KV, Petersen NJ, Hallman DM, et al. Day 1 multiple organ dysfunction syndrome is associated with poor functional outcome and mortality in the pediatric intensive care unit. Pediatr Crit Care Med. 2009; 10:562–570
15. Leteurtre S, Martinot A, Duhamel A, et al. Validation of the paediatric logistic organ dysfunction (PELOD) score: Prospective, observational, multicentre study. Lancet. 2003; 362:192–197
16. Johnston JA, Yi MS, Britto MT, et al. Importance of organ dysfunction in determining hospital outcomes in children. J Pediatr. 2004; 144:595–601
17. Goldstein B, Giroir B, Randolph A; International Consensus Conference on Pediatric Sepsis. International pediatric sepsis consensus conference: Definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med. 2005; 6:2–8
18. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)–a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42:377–381
19. Walter JK, Benneyworth BD, Housey M, et al. The factors associated with high-quality communication for critically ill children. Pediatrics. 2013; 131 (Suppl 1):S90–S95
20. Lavrakas P, (Ed): Cognitive interviewing. In: Encyclopedia of Survey Research Methods. Thousand Oaks, CA, Sage Publications, Inc., 2013
21. White DB, Angus DC, Shields A-M, et al. A randomized trial of a family-support intervention in intensive care units. N Engl J Med. 2018; 378:2365–2375
22. Ludmir J, Netzer G. Family-centered care in the intensive careunit-what does best practice tell us? Semin Respir Crit Care Med. 2019; 40:648–654
23. Wong EL, Yam CH, Cheung AW, et al. Barriers to effective discharge planning: A qualitative study investigating the perspectives of frontline healthcare professionals. BMC Health Serv Res. 2011; 11:242
24. Aiyer M, Kukreja S, Ibrahim-Ali W, et al. Discharge planning curricula in internal medicine residency programs: A national survey. South Med J. 2009; 102:795–799
25. Salvatori P, Mahoney P, Delottinville C. An interprofessional communication skills lab: A pilot project. Educ Health (Abingdon). 2006; 19:380–384
26. Jones S, Rantell K, Stevens K, et al.; United Kingdom Pediatric Intensive Care Outcome Study Group. Outcome at 6 months after admission for pediatric intensive care: A report of a national study of pediatric intensive care units in the United kingdom. Pediatrics. 2006; 118:2101–2108
27. Nelson LP, Lachman SE, Li SW, et al. The effects of family functioning on the development of posttraumatic stress in children and their parents following admission to the PICU. Pediatr Crit Care Med. 2019; 20:e208–e215
28. Abela KM, Wardell D, Rozmus C, et al. Impact of pediatric critical illness and injury on families: An updated systematic review. J Pediatr Nurs. 2020; 51:21–31
29. Milton A, Brück E, Schandl A, et al. Early psychological screening of intensive care unit survivors: A prospective cohort study. Crit Care. 2017; 21:273
30. Hynan MT, Steinberg Z, Baker L, et al. Recommendations for mental health professionals in the NICU. J Perinatol. 2015; 35 (Suppl 1):S14–S18
31. Omilion-Hodges LM, Swords NM. Communication matters: Exploring the intersection of family and practitioner end of life communication. Behav Sci (Basel). 2017; 7:E15
32. October TW, Dizon ZB, Arnold RM, et al. Characteristics of physician empathetic statements during pediatric intensive care conferences with family members: A qualitative study. JAMA Netw Open. 2018; 1:e180351
33. Trowbridge A, Bamat T, Griffis H, et al. Pediatric resident experience caring for children at the end of life in a children’s hospital. Acad Pediatr. 2020; 20:81–88