Priorities and Outcomes for Youth-Adult Transitions in Hospital Care: Perspectives of Inpatient Clinical Leaders at US Children’s Hospitals

We surveyed inpatient general pediatric service leaders from US children’s hospitals between January and July of 2016. The detailed survey development and distribution methods have been described previously.¹⁷  Briefly, hospitals were identified by using the online Children’s Hospital Association directory. We sought a single respondent from each hospital. To identify the leader of inpatient care with the most knowledge of his or her institution’s practices related to the transition to adult inpatient care, we used a structured protocol. This included the following steps: before sending the survey, study team members contacted each hospital’s inpatient general pediatric service leaders to describe the purpose of the study, confirm contact information, and when relevant, receive contact information for a designee having more expertise to receive the survey. We excluded hospitals without inpatient general pediatrics services (eg, rehabilitation or subspecialty-only facilities). The study was approved by the University of Wisconsin-Madison’s Institutional Review Board.

A novel survey instrument was developed from consensus-based recommendations of a multidisciplinary team of internal medicine and pediatric physicians, nurses, case managers and social workers, child life specialists, and patient and family advisors at our institution. Together, members of the team had experience working in >10 other health systems. The survey was refined after pilot testing with 5 pediatricians not involved in the study at 3 different institutions. Specific inpatient youth-adult transition activities and potential outcomes of this transition were developed from ideal-state process mapping and creation of a previously published conceptual framework of determinants for implementing a formal inpatient transition initiative.¹⁷  In total, 21 specific transition activities and 13 inpatient transition outcomes were included. The 21 transition activities were also grouped in categories by using the Six Core Elements framework¹² : policy, tracking and monitoring, readiness, planning, transfer of care, or transfer completion.

Respondents rated perceived importance and feasibility of performing each of the 21 transition activities and measuring each of the 13 transition outcomes using 5-point Likert scales (1 = “not at all,” 2 = “a little,” 3 = “somewhat,” 4 = “very,” or 5 = “extremely”). Respondents subsequently indicated whether these same activities were performed at their institution at least 50% of the time.

In addition, respondents were asked a series of questions about when, how, and for whom inpatient transition processes should take place. Using 5-point Likert scales, respondents rated the appropriateness of patient characteristics and other contextual triggers to identify the need for inpatient transition activities, as well as the best timing, age, and settings to initiate planning. The survey also contained questions on the respondents’ hospital characteristics, including administrative structure (freestanding or hospital within hospital), university-based or community, urban or rural, estimated number of discharges per week, scope of practice (inpatient only, combination inpatient and outpatient), presence or absence of specific hospital to receive adult patients, and whether ambulatory pediatric-adult transition processes existed.

The University of Wisconsin-Madison Survey Center conducted this survey using the Qualtrics Internet-based platform. An invitation with a printed URL to the survey was first sent to potential respondents by US mail. Up to 2 e-mail reminders containing the URL at 5 and 10 days after the initial mailing were then sent. One reminder phone call followed by a mailed paper copy of the survey questionnaire 2 weeks after the last emailed request was sent to nonresponders. Participants received $20 as an incentive.

Mean importance and feasibility ratings for each of the 21 transition activities were calculated. We also calculated a mean rating for the transition activities grouped by each of the Six Core Elements categories (eg, policy, tracking and monitoring, etc). After confirming ratings were normally distributed, linear regression was used to identify statistically significant associations between feasibility ratings and children’s hospital characteristics. Next, logistic regression was used to identify associations between current performance (or lack thereof) of each transition activity and importance or feasibility ratings. Lastly, inpatient transition outcomes were summarized with descriptive statistics, and importance and feasibility ratings were plotted on a 2-by-2 matrix to identify the outcomes perceived to be most important and feasible to measure. We considered the highest-priority outcomes to have both mean ratings above the median of all outcomes and to have the majority (>50%) of respondents rate them very or extremely important and feasible to measure. All analyses were completed in Stata SE version 14.0 (Stata Corp, College Station, TX), and P < .05 was considered statistically significant.

In total, 96 of 195 (49.2%) individuals responded (Table 1), with the most common respondent roles including division director (41%) or medical director (25%). Respondents were primarily from university-based centers (78%) in urban settings (88%). Children’s hospitals were freestanding (36%) or hospitals within hospitals (same building [43%]; separate building [22%]). Additional respondent and hospital characteristics are described elsewhere.¹⁷ 

Importance was rated higher than feasibility across the transition activities (Table 2). The most important activities (mean of 4.5 for each) were identifying patients needing or overdue for transition, discussing timing with youth and/or families, and informing youth and/or families that future stays would be at an adult facility. These same items had the highest feasibility ratings (means of 3.7–3.9).

When grouped by the Six Core Elements categories, the activities within tracking and monitoring had the highest importance (mean 4.5) and feasibility (mean 3.8) ratings (Table 3). Transfer completion activities, which included interactions during the first nonpediatric stay between pediatric providers and youth and/or family as well as child life, had the lowest perceived importance (mean 3.0) and feasibility (mean 3.1).

Certain hospital characteristics were associated with transition activity feasibility ratings (Table 4). Having a youth-adult transition policy that includes inpatient care had higher perceived feasibility for respondents from freestanding, university-based, and higher-volume children’s hospitals. Tracking and monitoring and transfer of care activities had higher perceived feasibility for respondents from higher-volume children’s hospitals. Transfer completion activities had higher perceived feasibility for respondents from hospitals within hospitals and those having a specific hospital to receive adult patients. Of note, however, having an ambulatory youth-adult transition process was not associated with perceptions of feasibility. Readiness and planning activity feasibility ratings were not associated with any hospital characteristics.

Relationships between current transition activity performance and importance or feasibility ratings are shown in Supplemental Figs 1 and 2. Importance ratings were inconsistent and weakly associated with performing any transition activities (Supplemental Fig 2). Higher feasibility ratings, on the other hand, were frequently associated with both performance of activities within that core element and also activities in other core elements (Supplemental Fig 3). For example, whereas higher feasibility ratings for planning were associated with providing a transition care plan to the patient and/or family (a specific planning activity), higher feasibility ratings for planning were also significantly associated with performing activities in each of the other core elements.

More than 80% of respondents reported that at least some of their patients need this type of transition process (Table 5), with nearly one-third (30%) estimating that most or all need it. Each of the queried patient characteristics (ie, medical or social complexity, past hospital use, or likelihood for future hospital use) was felt to be a very or extremely appropriate signal for the need for inpatient transition process by >75% of respondents. Initiating inpatient transition when general transition planning began was reported as very or extremely appropriate by 81% of respondents.

The mean ratings for 5 outcomes were above the median for both importance and feasibility to measure (Fig 1): (1) number of emergency department (ED) visits in early adulthood, (2) ambulatory provider satisfaction with transition, (3) patient and family satisfaction with overall transition, (4) provider confidence in patient and family managing their health and health care, and (5) patient and family confidence managing their health and health care. Among these 5 items, 2 (ED visits and patient and family satisfaction) were also rated very or extremely important and very or extremely feasible to measure by >50% of respondents (Table 6). More than 50% of respondents also rated inpatient provider satisfaction, number of hospitalizations, and average length of stay per visit as both very or extremely important and very or extremely feasible to measure outcomes of inpatient transition processes.

In this national survey, we characterized perceptions of inpatient general pediatric service leaders at US children’s hospitals about important and feasible youth-adult transition activities and outcomes relevant to general inpatient care. We sought to inform research and improvement priorities by highlighting activities perceived to be important and feasible. Initial priorities suggested in this study (to develop systems to proactively identify and communicate with youth and families needing inpatient youth-adult transition processes) parallel those from national youth-adult health care transition guidelines.^8,12  Inpatient provider opinions that these activities may be most appropriately accomplished during ambulatory care is consistent with research in which authors suggest that transfers to adult care are best conducted when the youth’s health is stable (ie, not hospitalized).¹⁹  According to the surveyed stakeholders, ED visits during early adulthood and patient and/or family experience with overall transition may be particularly useful outcomes because the majority considered them important and feasible to measure.

Although broad concepts (eg, assessing readiness) are considered important from both hospital-focused and comprehensive youth-adult transition persepctives,^9,16  with our research, we extend these concepts by suggesting a set of inpatient-specific activities that could theoretically be incorporated into existing transitional care strategies. For example, assessment of transition readiness may need to cover aspects of inpatient care for youth with high risk for future hospitalization. The Transition Readiness Assessment Questionnaire,^20,21  a tool commonly used to assess readiness, has a section on appointment keeping. For high-risk youth, one might conceive of a complementary set of questions that are focused on navigating hospital structures or managing postdischarge questions.

We were surprised to not find more relationships between institutional characteristics and perceptions of feasibility. It was striking that the feasibility of these inpatient transition activities was perceived similarly whether a respondent’s institution had an ambulatory youth-adult transition process. We speculate this is because many existing youth-adult transition processes may not have a connection to inpatient care. Moreover, readiness and planning, the 2 core elements with the largest numbers of transition activities had no associations between perceived feasibility and hospital characteristics. This may imply that hospitals different from each other could view these aspects of transition in relatively similar ways. Nevertheless, the success of implementation efforts will likely have strong contextual influences, and understanding unique barriers and facilitators to transition activity performance is important. For example, respondents from lower-volume hospitals did perceive less feasibility for policy, tracking and monitoring, and transfer activities. Implementation initiatives for these activities may require different strategies and resources on the basis of hospital size, structure, and scope.

Current performance of the surveyed transition activities was associated with perceived feasibility; however, the direction of the relationships remains unclear. These activities could be rated feasible because they were already being performed rather than being performed because they were perceived to be feasible. If participants only rated items as feasible when they were already being performed, one might have expected the relationships between feasibility and activity performance to be more limited to the core element being rated. However, importance perceptions were notably weak predictors of transition activity performance. Additional research should be used to clarify and address the barriers to performing transition activities that stakeholders consider to be “important.”

The outcomes of inpatient youth-adult transition rated most important and feasible to measure correspond to trends in the youth-adult transition field. For example, ED visits, hospital use, and patient experience have been relatively common outcomes in transition research and quality improvement.^{13–15,22,23}  Provider and patient confidence to manage health and health care (the second and third most important perceived outcomes overall in our study) are reminiscent of the transition field’s focus on readiness and self-management.^12,24–26  These measures allude to notions of general patient activation, which has been linked to a number of important health outcomes.^27,28  Understanding how youth-adult transitional care acts as a determinant (or potential intervention point) to increase general activation of CCOC during early adulthood would likely be a valuable focus for subsequent study.

Our study findings need to be interpreted with several limitations in mind. The survey provides inpatient clinical leader perspectives only, and other stakeholders may have different perspectives. The novel survey instrument has unknown reliability and validity, and although it was developed with input from a multidisciplinary adult and pediatric health care team including patients and parents, important transition activities and outcomes may be missing. Although we attempted to systematically identify the best respondent at each institution, it is possible that responses may have differed if another individual participated from each institution.

Despite these limitations, we anticipate that the relevance of this work will continue to grow as the prevalence of adults with CCOCs continues to increase. We suspect that surveyed inpatient general pediatric service leaders view these specific activities as important because they recognize that acute hospital care is distinct from other health care encounters, often having higher stakes. For example, hospitalization reflects times of crisis and can expose unique clinical, financial, and psychosocial vulnerabilities.

The results of this study contribute toward a roadmap for future prospective research. To sharpen the focus on inpatient youth-adult transition care, authors of subsequent studies should collect adult inpatient provider, patient, family caregiver, subspecialist, and ambulatory provider opinions to refine these activities and outcomes. Although inpatient providers felt that only certain patients may benefit from transition planning and that ambulatory clinic is an appropriate setting to conduct these activities, both perspectives require validation with other stakeholders. If ambulatory settings prove to be a viable context to introduce planning and preparation for future adult-oriented hospital care, it will be important to understand how providers with relevant knowledge (hospital-based physicians or nurses) can efficiently collaborate with outpatient providers to achieve this goal. Results of studies in which authors adapt and implement these elements within existing comprehensive transition programs at children’s hospitals would be informative. Researchers of subsequent work should also refine the appropriate populations, triggers, and settings for inpatient youth-adult transition care.