Priorities for Safer In-Person School for Children With Medical Complexity During COVID-19

We adapted the Child Health and Nutrition Research Initiative (CHNRI)²⁴  to identify the top priorities to support safer in-person school attendance for CMC across Wisconsin (Fig 1). This rapid, replicable and transparent method allows for crowdsourcing of ideas from a large group of diverse stakeholders and then engages technical experts in prioritizing these ideas using predefined criteria (eg, equity, sustainability).²⁵  The CHNRI method was originally developed to direct priority-setting for investments to improve child health and reduce inequities globally.²⁶  Used by the World Health Organization, this well-established method has been shown to be easily adaptable to various applications and has been used in over 50 priority-setting research initiatives,²⁷  including at least one during the COVID-19 pandemic.²⁸  When compared with other priority-setting methods, the CHNRI approach offers multiple advantages, including flexibility, transparency, active stakeholder engagement, and a systematic process and quantitative scoring with relatively low upfront resource commitment.^29,30  Our complete adapted CHNRI process is detailed in Supplemental Table 4. The University of Wisconsin’s Institutional Review Board determined this study to be exempt research.

This study was conducted in Wisconsin, which is comprised of 72 counties, 464 public school districts, and 792 private schools. As a “local control” state, each of Wisconsin’s public school district’s school boards have the statutory authority to decide its own instructional standards and curricula, provided the policies comply with the state constitution. As has been the case in many states across the United States, decisions around COVID-19 policies have been variable across Wisconsin school districts. To bridge the gap between public school districts and health care providers, districts partnered with medical advisors from the University of Wisconsin School of Medicine and Public Health in the fall of 2020. Given the diversity of policies across districts, the stakeholder groups and technical expert panel were selected to capture varying views from across the state.

In May 2021, we sought input from a broadly representative sample of families of CMC, family advocates, school staff, clinicians, administrators and policymakers across the state.³¹  These stakeholders were asked to submit their ideas for ways to achieve safe school attendance for CMC through a brief, anonymous, online survey (Research Electronic Data Capture, REDCap).^32,33  In this study, we defined CMC to stakeholders as “children having severe medical conditions who often rely on medical devices or people to help them with daily activities. They are fragile and at highest risk from COVID-19.” The survey was sent via e-mail and listserv distributions to individuals and groups identified by the study team as having an interest in CMC school attendance (Supplemental Table 5). The survey was developed by the study team and refined after piloting with 5 school staff and administrators, 4 parents of CMC, and 3 clinicians. It was offered in both English and Spanish.

The survey started with an open-ended prompt, “It is safe for children with complex health issues and those around them (families, teachers, classmates, etc.) to go to school in person during the COVID-19 pandemic if/when:” The prompt was designed to elicit a wide array of short, free-text responses. Instructions encouraged respondents to complete the prompt with only single ideas and to contribute as many separate responses to the prompt as they wished.

Respondents were then asked to consider 5 scoring criteria chosen from the CHNRI literature,³⁴  and defined as follows:

1. Equity: Can this idea be accessed by everyone who needs it, regardless of race, income, geography, etc.?

2. Impact on COVID-19: Does this idea have the potential to reduce the spread of COVID-19?

3. Practicality: Can this idea be done in a consistent way at most schools?

4. Sustainability: Can this idea be done throughout the pandemic and beyond, if needed?

5. Cost: Will the cost of this idea be justifiable?

Respondents scored each criterion's importance on a 5-point Likert scale (1: Not too important; 2: Somewhat important; 3: Moderately important; 4: Very important; and 5: Extremely important). The purpose of this step was to use these scores to calculate weights for the expert ratings.

Respondents also self-identified their role(s) (ie, parent/guardian/caregiver of CMC, teacher, administrator, other school personnel, clinician, policymaker, or other) and basic demographic information. Because of known racial and ethnic disparities in COVID-19 transmission and mortality, we included race and ethnicity to ensure participants represented a diverse population. Participants were given 10 days to complete the survey; responses were not linked to personal identifiers. Although an individual could be a member of more than one stakeholder organization, they were asked to complete the survey only once.

Three team members coded, edited, and combined crowdsourced ideas into a final list of unique ideas or “candidate priorities” through an iterative, consensus-seeking approach. Thematic analysis³⁵  was used to identify key themes that emerged during aggregation and synthesis. Study team members reviewed the final list of candidate priorities to ensure representation of all submitted ideas.

Leaders of the stakeholder groups (Supplemental Table 5) and the study team nominated 35 technical experts having sufficient expertise to judge the merit of each candidate priority using the 5 predefined scoring criteria. Experts included families of CMC, advocates, school personnel, clinicians, administrators, and policymakers. To ensure credibility and fairness of priority determinations, a wide range of individuals from diverse professional, demographic and geographic backgrounds was selected.^31,34  An initial online kickoff meeting provided an overview of project goals, candidate priorities, and scoring instructions. Experts were provided with a sample survey question, response category definitions, and an opportunity to ask questions.

To score candidate priorities, experts were e-mailed a link to an online survey (REDCap^32,33 ). They were asked to rate whether each candidate priority met each of the 5 scoring criteria: equity, impact on COVID-19, practicality, sustainability, and cost according to the definitions provided (Yes = 1, No = 0, and Unsure = 0.5). For example, if a priority was not likely to affect the spread of COVID-19, they were instructed to answer “No” for “Impact on COVID-19.” As with crowdsourcing, respondents also self-identified their role(s) and basic demographic information. Experts were given 2 weeks and nonresponders 2 reminders to complete the survey. They were offered $250 compensation for their participation.

Following CHNRI convention,²⁶  for each candidate priority we first calculated unweighted (intermediate) scores for each of the 5 criteria. For example, if 17 of 34 experts rated a candidate priority as “yes” for the equity criterion, 10 rated it as “unsure”, and 7 rated it as “no”, the intermediate score was calculated as {[(17 × 1) + (10 × 0.5) + (7 × 0)]/34} × 100, or 65%. The overall score was then calculated as the mean of all criteria scores.

Expert ratings were then weighted according to the relative importance of each criterion.²⁶  In this way, the values of the stakeholder community were further reflected within the prioritization results.^31,34  Using the scores gathered during crowdsourcing, weights for each criterion were calculated as a criterion’s mean importance score (from 1–5) divided by the sum of all criteria mean importance scores. A weighted overall score was calculated as the sum of each intermediate criterion score multiplied by its criterion weight.

Leaders of the stakeholder groups (Supplemental Table 5) and the study team set a priori thresholds, ie, minimum acceptable scores (from 0–100) for each criterion for an item to be considered a priority. Assigned thresholds were 75 for equity, 65 for impact on COVID-19, 65 for practicality, 50 for sustainability, and 30 for cost. For a candidate priority to be included in the final list of consensus priorities, all relevant criterion scores had to be above the assigned thresholds.

Two team members independently identified whether priorities were specific to CMC (vs all children) and recommended in the CDC’s Guidance for COVID-19 Prevention in K-12 Schools,²⁰  and AAP’s Caring for Children and Youth with Special Healthcare Needs during the COVID-19 Pandemic²¹  and COVID Guidance for Safe Schools²²  to highlight areas where our results were complementary and unique. Discrepancies were resolved through group discussion. The final consensus priorities were defined as those being specific to CMC.

The expert group was reconvened for a virtual concluding meeting to refine and validate findings. The group also discussed strategies to disseminate and implement the priorities across Wisconsin schools and communities. Family (B.K., R.B.), school (L.W.), clinician (M.E., S.B.) and administrator (G.M., D.P.) participants were coauthors to ensure that the interpretation and presentation of consensus priorities was consistent and balanced from these perspectives.

In total, 460 stakeholders responded to the crowdsourcing survey (Table 1). Nearly one-third were parents of a CMC (32%) and identified with going to school (34%) or working (36%) in a rural area. Approximately one-half (48%) represented teachers or other school personnel (27%), clinicians (12%), and administrators (9%).

The survey generated 1166 ideas with a mean of 2.6 ideas per respondent. The study team aggregated and synthesized these ideas into a list of 87 unique candidate priorities within 10 themes: (1) COVID-19 screening, exposure and testing; (2) masking, (3) school environment and cleaning, (4) staffing and caregiving, (5) curriculum and flexibility, (6) health risks and planning, (7) school protocols, (8) transportation, (9) COVID-19 vaccination, and (10) community.

Thirty-four technical experts completed the scoring survey (response rate 97%, Table 1). Experts represented parents of CMC (38%), teachers or other school personnel (47%), and clinicians (18%).

Experts scored the 87 unique candidate priorities according to the scoring criteria. The intermediate and overall scores for each candidate priority are shown in Supplemental Fig 2. Scores illustrate the varying degrees of support for each candidate priority according to perceived equity, impact, practicality, sustainability, and cost. The highest scored priorities focused on educating staff and families about the unique health risks of contracting COVID-19 for CMC and developing plans that outline a CMC’s daily routines and contingency plans that account for COVID-19 safety during IEP meetings. The lowest scored priorities were often very isolating for CMC (eg, taught one-on-one, in a separate location, afterschool hours, or does not share transportation), or deemed impractical (eg, daily testing, single staff provides all interactions for a child for teaching, care routines, or mobility).

Many of the top 35 priorities were referenced within guidelines subsequently published by either the CDC or AAP and more than half were relevant to all children at school and not specific to CMC (Table 2). The 10 consensus priorities that targeted CMC specifically were synthesized and are summarized in Table 3. Expert ratings also reinforced general COVID-19 mitigation strategies recommended by the CDC²⁰  and AAP,^21,22  including promoting vaccination for all eligible students, families, and school staff; facilitating/practicing cleaning and disinfecting practices, adequate ventilation, hand-washing, and respiratory etiquette, and participating in screening, contact tracing, isolation, and quarantine as indicated.

We adapted a rapid, transparent method with a broad and diverse group of statewide stakeholders to elicit and prioritize ideas for safer in-person school for CMC during the COVID-19 pandemic. This systematic process led to the development of a consensus list of priorities deemed to be equitable, practical, sustainable, and cost-effective. These findings simultaneously reinforce CDC²⁰  and AAP^21,22  guidelines and introduce additional unique and practical ways to support safer school attendance for CMC, their classmates, and school staff. Our results may be relevant to those in other states, and this study also serves as a template that others could similarly rapidly generate findings that reflect the views stakeholders in their distinct local context.

Our findings provide validation for CDC²⁰  and AAP^21,22  recommendations from the perspective of families of CMC and school staff. Our results also underscore that CMC needs for safe school attendance are likely not yet fully met. In fact, when considering priorities to support safer CMC school attendance, equity was given the greatest importance, and ideas that segregated CMC from other students at school, which are likely not compliant with federal antidiscrimination legislation, were rated at the bottom of the priority list. Over the 2021–-2022 school year, our team will be monitoring CMC parent perceptions and experiences with school mitigation strategies to determine whether changes occur over time, and whether more families are opting to send their CMC to in-person school. Ongoing research will be needed to determine whether these priorities are effective at promoting attendance while keeping CMC, families, and communities safe.

An important contribution of this study is that, by inclusively engaging with statewide stakeholders throughout this process, we anticipate the results will have a higher degree of relevance and face validity than recommendations developed and disseminated through more traditional or “top-down” modes.³⁶  The results have already provided a framework upon which we are now developing publicly available tools to support implementation. Our dissemination web portal³⁷  (www.reset4kids.org) contains downloadable content including a summary of the priorities, a family guide, videos to educate staff and families on the increased health risks of COVID-19 for CMC, sample letter templates for health care providers to schools, vaccination information, and more.

Although we cannot yet assess whether our approach will translate into higher rates of uptake of the consensus priorities, we do have early evidence suggesting the materials resulting from this work are being spread. Our dissemination portal³⁷  generated >2300 views, including >425 downloads in the first 4 weeks. Our content was also included in the Wisconsin Department of Public Instruction 2021–2022 COVID-19 Infection Control and Mitigation Measures for Wisconsin Schools document.³⁸  Additional tools will be developed on an ongoing basis in response to feedback from families and schools as the school year begins.

Although the focus of this work is on CMC, it is important to note that most of the top priorities have relevance to all children. We interpret this observation as reinforcement that the safety of CMC relies upon practices that promote the safety of all students and staff at school. For example, as schools consider policy decisions regarding mask use, they must consider the influence that this broad policy decision has on children with high-risk medical conditions that are particularly vulnerable to COVID-19. CMC are also entitled to reasonable accommodations under Title II of the Americans with Disabilities Act and Section 504 of the Rehabilitation Act. A universal mask policy may be necessary if that policy facilitates full participation of CMC in the educational programming that would otherwise be inaccesible.³⁹  Our findings also reiterate the idea that schools must continue to focus on a multilayered approach to prevent COVID-19 transmission.

This study has important limitations. Ideas were elicited from and prioritized by stakeholders across one state and some stakeholders were underrepresented with respect to census demographics (eg, non-Hispanic Black respondents); findings may or may not be generalizable to other populations. Crowdsourcing survey response rates were deliberately not estimated given that the survey link was sent to multiple listservs of which may include many of the same potential respondents and their responses were anonymous. This should not affect the validity of findings, as the goal was to achieve saturation of ideas from a large and heterogeneous group of stakeholders rather than determine frequencies of responses. The CHNRI method relies on technical experts having the necessary knowledge to make valid judgements on each of the rating criteria, which we were unable to verify. Similarly, we do not yet know whether the priorities are effective at preventing COVID-19 spread or ensuring school safety. Finally, we recognize that these priorities will change over time based on new knowledge, infection patterns, and policies.

Despite these limitations, this study provides a roadmap for other school communities, clinicians, and academic centers to replicate. Elicited priorities validate and build upon CDC and AAP recommendations and offer a pragmatic guide for families, school staff, clinicians, and policymakers for decision-making and resource deployment to achieve safer school attendance for CMC.

AAP

American Academy of Pediatrics

CDC

Centers for Disease Control and Prevention

CHNRI

Child Health and Nutrition Research Initiative

CMC

children with medical complexity

IEP

individualized education plan

PPE

personal protective equipment