Advertisement

Autism Spectrum Disorder Screening During the COVID-19 Pandemic in a Large Primary Care Network

  • Kate E. Wallis
    Correspondence
    Address correspondence to Kate E. Wallis, MD, MPH, Roberts Center for Pediatric Research, Room 5291, 2716 South St, Philadelphia, PA 19146
    Affiliations
    Division of Developmental and Behavioral Pediatrics (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Autism Research (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Ekaterina Nekrasova
    Affiliations
    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    The Possibilities Project (E Nekrasova, AG Fiks, BP Jenssen), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Amanda E. Bennett
    Affiliations
    Division of Developmental and Behavioral Pediatrics (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Autism Research (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Alexander G. Fiks
    Affiliations
    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    The Possibilities Project (E Nekrasova, AG Fiks, BP Jenssen), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Marsha Gerdes
    Affiliations
    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Brian P. Jenssen
    Affiliations
    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    The Possibilities Project (E Nekrasova, AG Fiks, BP Jenssen), Children's Hospital of Philadelphia, Philadelphia, Pa
    Search for articles by this author
  • Judith S. Miller
    Affiliations
    Division of Developmental and Behavioral Pediatrics (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Autism Research (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Child and Adolescent Psychiatry and Behavioral Sciences (JS Miller and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Psychiatry, Perelman School of Medicine (JS Miller and W Guthrie), University of Pennsylvania, Philadelphia, Pa
    Search for articles by this author
  • Di Shu
    Affiliations
    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Biostatistics, Epidemiology and Informatics (D Shu), University of Pennsylvania Perelman School of Medicine, Philadelphia, Pa
    Search for articles by this author
  • Whitney Guthrie
    Affiliations
    Division of Developmental and Behavioral Pediatrics (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Pediatrics (KE Wallis, AE Bennett, AG Fiks, M Gerdes, BP Jenssen, JS Miller, D Shu, and W Guthrie), Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pa

    Center for Autism Research (KE Wallis, AE Bennett, JS Miller, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Center for Pediatric Clinical Effectiveness (KE Wallis, E Nekrasova, AG Fiks, M Gerdes, BP Jenssen, D Shu, and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Child and Adolescent Psychiatry and Behavioral Sciences (JS Miller and W Guthrie), Children's Hospital of Philadelphia, Philadelphia, Pa

    Department of Psychiatry, Perelman School of Medicine (JS Miller and W Guthrie), University of Pennsylvania, Philadelphia, Pa
    Search for articles by this author
Published:April 20, 2022DOI:https://doi.org/10.1016/j.acap.2022.04.005

      Abstract

      Objective

      To assess the impact of the COVID-19 pandemic on screening for autism spectrum disorder (ASD) and screening equity among eligible children presenting for well-child care in a large primary care pediatric network, we compared rates of ASD screening completion and positivity during the pandemic to the year prior, stratified by sociodemographic factors.

      Methods

      Patients who presented for in-person well-child care at 16 to 26 months between March 1, 2020 and February 28, 2021 (COVID-19 cohort, n = 24,549) were compared to those who presented between March 1, 2019 and February 29, 2020 (pre-COVID-19 cohort, n = 26,779). Demographics and rates of completion and positivity of the Modified Checklist for Autism in Toddlers with Follow-up (M-CHAT/F) were calculated from the electronic health record and compared by cohort using logistic regression models.

      Results

      Total eligible visits decreased by 8.3% between cohorts, with a greater decline in Black and publicly insured children. In the pre-COVID-19 cohort, 89.0% of eligible children were screened at least once, compared to 86.4% during the pandemic (P < 0.001). Significant declines in screening completion were observed across all sociodemographic groups except among Asian children, with the sharpest declines among non-Hispanic White children. Sociodemographic differences were not observed in screen-positive rates by cohort.

      Conclusions

      Well-child visits and ASD screenings declined across groups, but with different patterns by race and ethnicity during the COVID-19 pandemic. Findings regarding screen-completion rates should not be interpreted as a decline in screening disparities, given differences in who presented for care. Strategies for catch-up screening for all children should be considered.

      Keywords

      What's New
      Well-child visits for young children and rates of autism-screening completion across a primary care pediatric network declined during the COVID-19 pandemic. Racial and ethnic differences in screening completion persist, but disparities narrowed slightly due to differences in well-child care receipt.
      During the Coronavirus Disease of 2019 (COVID-19) pandemic, provision of well-child care declined in response to policies to limit infection

      Pediatrician life and career experience study (PLACES) Special Issue. In: AAP, ed. Going Places Newsletter June 2020.

      and parental concern about seeking non-urgent health care.
      • Ciacchini B.
      • Tonioli F.
      • Marciano C.
      • et al.
      Reluctance to seek pediatric care during the COVID-19 pandemic and the risks of delayed diagnosis.
      Ongoing surveillance and universal screening for autism spectrum disorder (ASD) at 18- and 24-month visits are recommended to promote early detection and initiation of services.
      • Hyman S.L.
      • Levy S.E.
      • Myers S.M.
      Identification, evaluation, and management of children with autism spectrum disorder.
      Prior to the pandemic, rates of ASD screening across the Children's Hospital of Philadelphia (CHOP) Care Network were high, with 91% of children screened at least once between 2011 and 2015. However, disparities in screening completion existed by race and ethnicity, family language, and socioeconomic status. Lower rates of having at least one screening were observed among children of Black (83.4%), Asian (91.6%), or multi-racial groups (91.8%) compared to White children (96.8%); those from lower income families (85.1%) compared to higher-income families (96.8%); those with public insurance (85.7%) compared to private (95.4%); and those who spoke Spanish at home (83.6%) compared to English only (91.2%).
      • Guthrie W.
      • Wallis K.
      • Bennett A.
      • et al.
      Accuracy of autism screening in a large pediatric network.
      Disparities in screening at both 18 and 24 months were even larger, in part because of differences in visit attendance.
      • Guthrie W.
      • Wallis K.
      • Bennett A.
      • et al.
      Accuracy of autism screening in a large pediatric network.
      The COVID-19 pandemic also brought about and raised awareness of perpetual racial, ethnic, and socioeconomic disparities for children with respect to health care access and quality and patient outcomes, including disease incidence and death.
      • Dennis-Heyward E.A.
      • Shah S.N.
      Pediatric COVID-19 disparities and prioritizing equity—children are not spared.
      While primary care offices in this study continued to prioritize well-child care visits for children under age 2 years and for those with missing vaccinations during the stay-at-home orders in March through May of 2020 consistent with Ameriacn Academy of Pediatrics guidance, care for older children was often deferred early in the pandemic. One hospital system found that children who are Black or Hispanic, younger, and privately insured were more likely to present for care than other groups in 2020.
      • Brown C.L.
      • Montez K.
      • Amati J.B.
      • et al.
      Impact of COVID-19 on pediatric primary care visits at four academic institutions in the Carolinas.
      However, the full impact of the COVID-19 pandemic on primary care provision and equity, including primary care's central role in ASD screening is incompletely understood.
      We sought to examine the impact of COVID-19 on ASD screening in the CHOP Network by enumerating and comparing rates of 1) ASD screening completion for eligible children presenting for well-child care and 2) positive screens. The rates during the pandemic were each compared to baseline metrics from the year prior, stratified by sociodemographic factors.

      Methods

      Patients who presented for in-person well-child care at 16 to 26 months of age (eligible screening ages, around the recommended 18 and 24-month visits) in the CHOP Care Network between March 1, 2020 and February 28, 2021 (COVID-19 cohort, n = 24,549) were compared to those who presented between March 1, 2019 and February 29, 2020 (pre-COVID-19 cohort, n = 26,779). The CHOP Care Network includes 29 sites in Pennsylvania and New Jersey that screen universally for ASD, using the Modified Checklist for Autism in Toddlers with Follow-up (M-CHAT/F).
      • Robins D.L.
      • Fein D.
      • Barton M.L.
      • et al.
      The Modified Checklist for Autism in Toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders.
      ,
      • Robins D.L.
      • Casagrande K.
      • Barton M.
      • et al.
      Validation of the modified checklist for Autism in toddlers, revised with follow-up (M-CHAT-R/F).
      The M-CHAT/F, rather than the M-CHAT–R/F, is used across the Network, as the accuracy of these 2 versions is comparable
      • Robins D.L.
      • Casagrande K.
      • Barton M.
      • et al.
      Validation of the modified checklist for Autism in toddlers, revised with follow-up (M-CHAT-R/F).
      and versions are often used interchangeably.
      • Khowaja M.K.
      • Hazzard A.P.
      • Robins D.L.
      Sociodemographic barriers to early detection of autism: screening and evaluation using the M-CHAT, M-CHAT-R, and Follow-Up.
      Families are able to complete the screen electronically prior to the visit through an electronic patient portal, or on a tablet in the waiting room before the visit, with results auto-populating in the electronic health record (EHR) for provider review.
      • Guthrie W.
      • Wallis K.
      • Bennett A.
      • et al.
      Accuracy of autism screening in a large pediatric network.
      Screening completion and results were extracted electronically from the EHR.
      Sociodemographic data (sex, race, ethnicity, insurance type, and preferred language) were also gathered from the EHR in order to identify disparities in care. Race and ethnicity were categorized into the following groups: Asian; Black, non-Hispanic (hereafter referred to as “Black”); Hispanic or Latino; Other (which included multiple races); and White, non-Hispanic (hereafter referred to as “White”). Preferred language was classified as English or Spanish. Families who did not speak English or Spanish (n = 1137) were excluded, as the screen is not yet available in other languages in the CHOP EHR. Primary insurance was classified as public (eg, Medicaid) or private.
      Chi square tests were used to compare the sociodemographic characteristics of children presenting for well child visits between the 2 cohorts. Univariate logistic regressions were used to estimate odds ratios of screening completion and a positive screen comparing 2 cohorts, within each sociodemographic group. Multivariate logistic regression models were used to evaluate main effects for each sociodemographic characteristic, cohort, and an interaction between sociodemographic characteristics and cohort, adjusting for other potential confounders. The pre-COVID-19 cohort was the reference group for all cohort comparisons. Analyses were conducted using Stata, version 15 (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC). This research was deemed exempt by CHOP's Institutional Review Board.

      Results

      Visit Counts

      During the pre-COVID period, 26,779 well-child visits were completed among children eligible for ASD screening, compared to 24,549 during the COVID period, a decrease of 8.3% (Table 1). Racial composition of children presenting for care varied between cohorts (P < 0.001); the proportion of Black children decreased between the pre-COVID and COVID cohorts, while proportions of White Hispanic children, and “Other” race children increased. Children presenting for care during the pandemic were slightly more likely to be privately insured than pre-pandemic (P = 0.048). Cohorts did not differ with respect to age or other sociodemographic factors.
      Table 1Demographic characteristics among eligible patients presenting for well-child care pre-COVID-19 and during COVID-19
      Total Eligible Visits (% of Total Annual Visits)
      Pre-COVID- 19 Cohort
      Pre-COVID- 19 Cohort (March 1, 2019 and February 29, 2020), Median age: 20.5 months, Total Eligible Visits = 26,779.
      COVID- 19 Cohort
      COVID- 19 Cohort (March 1, 2020 and February 28, 2021), Median age: 20.3 months, Total Eligible Visits = 24,549.
      P-value
      Total26,77924,549N/A
      Sex

      Female

      Male


      12798 (47.8%)

      13981 (52.2%)


      11929 (48.6%)

      12620 (51.4%)
      0.07
      Race and ethnicity

      Asian

      Black, Non-Hispanic

      Hispanic or Latino

      Other

      White, Non-Hispanic


      1194 (4.5%)

      5870 (21.9%)

      2442 (9.1%)

      4046 (15.1%)

      13227 (49.4%)


      1122 (4.6%)

      4872 (19.8%)

      2324 (9.5%)

      3864 (15.7%)

      12376 (50.4%)
      <0.001
      Statistically significant differences between proportions in COVID-19 cohort compared to pre-COVID-19 cohort.


      Preferred language

      English

      Spanish


      26150 (97.6%)

      627 (2.4%)


      23976 (97.7%)

      573 (2.3%)
      0.91
      Insurance

      Private

      Public


      17735 (66.2%)

      9044 (33.8%)


      16460 (67.1%)

      8089 (32.9%)
      0.048
      a Pre-COVID- 19 Cohort (March 1, 2019 and February 29, 2020), Median age: 20.5 months, Total Eligible Visits = 26,779.
      b COVID- 19 Cohort (March 1, 2020 and February 28, 2021), Median age: 20.3 months, Total Eligible Visits = 24,549.
      low asterisk Statistically significant differences between proportions in COVID-19 cohort compared to pre-COVID-19 cohort.

      ASD Screening Completion

      For screening completion rates (Table 2), 89.0% of eligible children who presented for a well-child visit completed the M-CHAT/F screening pre-pandemic, compared to 86.4% during the pandemic (P < 0.001). In both cohorts, White, English-speaking, and privately insured children had higher rates of screening completion compared to all other racial and ethnic groups, Spanish-speaking, and publicly insured children, respectively. Rates of screening completion were significantly lower in the COVID-19 cohort than the pre-COVID cohort across groups, with statistically significant decreases for both sexes, all racial groups except for Asian children, children from English- and Spanish-speaking families, and children with both types of insurance. Multivariate logistic models produced significant main effects of both race and ethnicity and cohort on screening completion (Table 3). Their interaction terms were significant for Asian children (aOR, 1.37, 95% CI 1.07–1.76), Black children (aOR 1.17, 95% CI 1.01–1.37), and children of “Other” racial groups (aOR 1.19, 95% CI 1.01–1.39). Overall, results indicate that 1) fewer Black children obtained well-child care during COVID-19 compared to other groups; however, 2) the decrease in screening completion during the pandemic was more pronounced among White children compared to Asian, Black, and “Other” race children, and thus 3) differences in screening completion by race and ethnicity were less pronounced during COVID-19 than before COVID-19. Other sociodemographic differences and interaction terms were not found statistically significant. To examine these trends, monthly rates in screening were visualized for both cohorts (Figure). In the pre-COVID-19 cohort, monthly screen-completion rates ranged from 86.5% to 92.7%, and in the COVID-19 cohort, from 76.0% to 90.8%. Noticeable discrepancies between 2 cohorts were seen before the reintroduction of tablets in waiting rooms.
      Table 2Univariate logistic regression analyses for M-CHAT/F
      M-CHAT/F indicates Modified Checklist for Autism in Toddlers with Follow-up.
      completion and M-CHAT/F positivity in the COVID-19 Cohort, compared to the pre-COVID-19 cohort in each sociodemographic group
      M-CHAT/F CompletePre-COVID- 19 Cohort
      Pre-COVID- 19 Cohort (March 1, 2019 and February 29, 2020), Median age: 20.5 months, Total Eligible Visits = 26,779.
      M-CHAT/F CompleteCOVID- 19 Cohort
      COVID- 19 Cohort (March 1, 2020 and February 28, 2021), Median age: 20.3 months, Total Eligible Visits = 24,549.
      Unadjusted Odds Ratio for M-CHAT/F Completion comparing two cohorts (95% CI)P-ValueM-CHAT/F PositivePre-COVID-19 Cohort
      M-CHAT/F indicates Modified Checklist for Autism in Toddlers with Follow-up.
      M-CHAT/F PositiveCOVID-19 Cohort
      Pre-COVID- 19 Cohort (March 1, 2019 and February 29, 2020), Median age: 20.5 months, Total Eligible Visits = 26,779.
      Unadjusted Odds Ratio for M-CHAT/F Positivity comparing two cohorts (95% CI)P-Value
      Total23,836/26,779 (89.0%)21,200/24,549 (86.4%)0.78 (0.74–0.82)<0.00011,855/23,836 (6.9%)1,705/21,200 (7.0%)0.14 (0.971.11)0.22
      Sex

      Female

      Male


      11406/12798 (89.1%)

      12430/13981 (88.9%)


      10286/11929 (86.2%)

      10914/12620 (86.5%)


      0.76 (0.71–0.83)

      0.80 (0.74–0.86)


      <0.0001

      <0.0001


      692/11406 (6.1%)

      1163/12430 (9.4%)


      666/10286 (6.5%)

      1039/10914 (9.5%)


      1.07 (0.961.20)

      1.02 (0.931.11)


      0.22

      0.67
      Race and Ethnicity

      Asian

      Black, Non-Hispanic

      Hispanic/Latino

      Other

      White, Non-Hispanic


      1023/1194 (85.7%)

      4946/5870 (84.3%)

      2131/2442 (87.3%)

      3578/4046 (88.4%)

      12158/13227 (91.9%)


      957/1122 (85.3%)

      3989/4872 (81.9%)

      1913/2324 (82.3%)

      3344/3864 (86.5%)

      10997/12376 (88.9%)


      0.97 (0.771.22)

      0.84 (0.76–0.93)

      0.68 (0.58–0.80)

      0.84 (0.74–0.96)

      0.71 (0.65–0.77)


      0.79

      0.001

      <0.001

      0.01

      <0.0001


      111/1023 (10.9%)

      711/4946 (14.4%)

      287/2131 (13.5%)

      271/3578 (7.6%)

      474/12158 (3.9%)


      103/957 (10.8%)

      646/3989 (16.2%)

      243/1913 (12.7%)

      271/3344 (8.1%)

      442/10997 (4.0%)


      0.99 (0.751.32)

      1.15 (1.021.29)

      0.94 (0.781.12)

      1.08 (0.901.28)

      0.13 (0.911.18)


      0.95

      0.02

      0.47

      0.41

      0.64
      Preferred Language

      English

      Spanish


      23324/26150 (89.2%)

      512/627 (81.4%)


      20786/23976 (86.7%)

      414/573 (72.3%)


      0.79 (0.75–0.83)

      0.60 (0.45–0.78)


      <0.001

      0.0002


      1761/23324 (7.6%)

      94/512 (18.4%)


      1635/20786 (7.9%)

      70/414 (16.9%)


      1.05 (0.981.12)

      0.91 (0.641.27)


      0.21

      0.57
      Insurance

      Private

      Public


      16043/17735 (90.5%)

      7793/9044 (86.2%)


      14454/16460 (87.8%)

      6746/8089 (83.4%)


      0.76 (0.71–0.81)

      0.81 (0.74–0.88)


      <0.0001

      <0.0001


      691/16043 (4.3%)

      1164/7792 (14.9%)


      643/14454 (4.5%)

      1062/6746 (15.7%)


      1.03 (0.921.15)

      1.06 (0.971.17)


      0.55

      0.18
      Bolded values indicate statistically significant differences between proportions in COVID-19 cohort compared to pre-COVID-19 cohort.
      a M-CHAT/F indicates Modified Checklist for Autism in Toddlers with Follow-up.
      b Pre-COVID- 19 Cohort (March 1, 2019 and February 29, 2020), Median age: 20.5 months, Total Eligible Visits = 26,779.
      c COVID- 19 Cohort (March 1, 2020 and February 28, 2021), Median age: 20.3 months, Total Eligible Visits = 24,549.
      Table 3Multivariate logistic regression analyses for M-CHAT/F
      M-CHAT/F indicates Modified Checklist for Autism in Toddlers with Follow-up.
      completion and M-CHAT/F positivity adjusted for demographic factors, COVID-19 Cohort,
      The COVID-19 Cohort (eligible children presenting for well-child care between March 1, 2020 and February 28, 2021) was compared to the Pre-COVID-19 Cohort (children presenting for well-child care between March 1, 2019 and February 29, 2020; reference group).
      as well as interaction terms between demographics and cohort
      M-CHAT/F CompletionM-CHAT/F Positivity
      Adjusted Odds Ratio(95% Confidence Intervals, CI)P-ValueAdjusted Odds Ratio(95% CI)P-Value
      Cohort

      Pre-COVID-19 Cohort

      COVID-19 Cohort


      Reference

      0.69 (0.62–0.76)


      Reference

      <0.0001


      Reference

      1.08 (0.91-1.27)


      Reference

      0.38
      Sex

      Female

      Male


      Reference

      0.98 (0.911.06)


      Reference

      0.62


      Reference

      1.63 (1.28-1.80)


      Reference

      <0.0001
      Interaction between chort and sex

      Female*COVID-19 cohort

      Male*COVID-19 cohort


      Reference

      1.05 (0.941.16)


      Reference

      0.40


      Reference

      0.94 (0.82-1.09)


      Reference

      0.44
      Race and ethnicity

      Asian

      Black, Non-Hispanic

      Hispanic or Latino

      Other

      White, Non-Hispanic


      0.53 (0.45–0.63)

      0.51 (0.45–0.56)

      0.76 (0.5–0.89)

      0.69 (0.61–0.77)

      Reference


      <0.0001

      <0.0001

      <0.001

      <0.0001

      Reference


      2.81 (2.25–3.50)

      2.33 (2.03–2.67)

      2.28 (1.90–2.74)

      1.66 (1.42–1.94)

      Reference


      <0.0001

      <0.0001

      <0.0001

      <0.0001

      Reference
      Interaction between Cohort and Race and Ethnicity

      Asian*COVID-19 cohort

      Black, Non-Hispanic*COVID-19 cohort

      Hispanic or Latino*COVID-19 cohort

      Other *COVID-19 cohort

      White, Non-Hispanic*COVID-19 cohort


      1.37 (1.07–1.76)

      1.17 (1.01–1.37)

      1.00 (0.801.24)

      1.19 (1.01–1.39)

      Reference


      0.01

      0.04

      1.00

      0.03

      Reference


      0.95 (0.921.35)

      1.11 (0.911.35)

      0.91 (0.701.19)

      1.03 (0.821.29)

      Reference


      0.76

      0.31

      0.50

      0.81

      Reference
      Preferred language

      English

      Spanish


      Reference

      0.56 (0.44–0.72)


      Reference

      <0.0001


      Reference

      1.19 (0.901.56)


      Reference

      0.22
      Interaction between cohort and preferred language

      English*COVID-19 cohort

      Spanish*COVID-19 cohort


      Reference

      0.82 (0.591.15)


      Reference

      0.25


      Reference

      0.92 (0.611.38)


      Reference

      0.69
      Insurance

      Private

      Public


      Reference

      0.88 (0.81–0.97)


      Reference

      0.008


      Reference

      2.78 (2.48–3.12)


      Reference

      <0.0001
      Interaction between cohort and insurance

      Private*COVID-19 cohort

      Public*COVID-19 cohort


      Reference

      1.03 (0.911.17)


      Reference

      0.64


      Reference

      1.00 (0.851.18)


      Reference

      0.98
      Bolded values indicate statistically significant differences between proportions in COVID-19 cohort compared to pre-COVID-19 cohort.
      a M-CHAT/F indicates Modified Checklist for Autism in Toddlers with Follow-up.
      b The COVID-19 Cohort (eligible children presenting for well-child care between March 1, 2020 and February 28, 2021) was compared to the Pre-COVID-19 Cohort (children presenting for well-child care between March 1, 2019 and February 29, 2020; reference group).
      Figure
      FigureMonthly screening rates in the Pre-COVID-19 cohort (children presenting for well-child care between March 1, 2019 and February 29, 2020) and the COVID-19 Cohort (eligible children presenting for well-child care between March 1, 2020 and February 28, 2021), with important policy changes as noted.

      ASD Screening Positivity

      In both cohorts, White, English-speaking, and privately insured children had lower rates of screening positivity compared to all other racial and ethnic groups, Spanish-speaking, and publicly insured children, respectively. There were no statistically significant differences in M-CHAT positivity between cohorts.

      Discussion

      During the COVID-19 pandemic, the number of clinicians, practices, and overall patient population in the network remained stable. However, well-child visits declined slightly, with the most pronounced declines among Black and publicly insured children. Though high proportions (>75%) of all children were screened before and during the pandemic, screening completion rates fell during the COVID-19 pandemic in every racial and ethnic group except in Asian children (for whom screening decreases were not statistically significant, perhaps as a result of smaller sample size of this group), with the most pronounced screening declines occurring among White children. This brought the overall screening completion rates across racial and ethnic groups closer to each other compared to pre-COVID-19. However, this should not be interpreted as a decrease in health disparities. Rather, it suggests that many children missed ASD screening, either because well-child care did not happen (eg, for a greater proportion of Black children), or because the ASD-screening portion of well-child care did not occur. Declines in well-child care for Black children during the pandemic are especially important to highlight and address, since other important components of well-child care beyond ASD screening may have also been missed. Black children experience both historic and contemporaneous barriers to high-quality health care,
      • Trent M.
      • Dooley D.G.
      • Dougé J.
      The impact of racism on child and adolescent health.
      and also suffered disproportionately from COVID-19 in multiple ways.
      • Van Dyke M.E.
      • Mendoza M.C.B.
      • Li W.
      • et al.
      Racial and ethnic disparities in COVID-19 incidence by age, sex, and period among persons aged <25 years - 16 U.S. Jurisdictions, January 1-December 31, 2020.
      These data add pressure to finding solutions to remedy both long-standing systemic injustices in health care receipt and the recent impacts of COVID-19.
      As all screening was completed electronically, decreased use of tablets for electronic screening in the waiting room during the height of the pandemic might have contributed to these reductions in screening completion (Figure). Tablet use decreased as an infection-control policy introduced to reduce the sharing of equipment between patients and time spent in shared waiting rooms, particularly early in the pandemic when relatively less was known about methods of transmission. Tablet use for screening completion was resumed across sites in June 2020. Study practices saw increasing activation and reduced disparities in virtual patient portal access throughout the course of the pandemic (through which screening was available)
      • Craig S.
      • Shen A.K.
      • Wallis K.
      • et al.
      How health systems can help address language barriers to achieve digital health equity.
      ; these changes likely contributed to increases in screening completion by July 2020.
      Competing personal priorities and additional responsibilities during the pandemic may have led to reductions in screening completion by caregivers (eg, families may have chosen to focus on concerns they considered more urgent). Alternative strategies to promote screening completion by families should be considered. These might include use of reminders, such as text messages, to complete screening that are not linked to the patient portal or reminding families to complete the screen on a personal device while in the office. A population-based approach might also be considered, which could entail sending out screening questionnaires to all patients at 18 and 24 months, even without a scheduled visit, and prioritizing scheduling patients for visits who screen positive. These approaches may be particularly beneficial in situations that might otherwise limit screening, such as a pandemic.
      The implications for these reduced rates of well-child care provision and ASD screening may be far-reaching if not acted upon. ASD-related concerns will need attention at future visits to identify children who were not screened and thus are at elevated risk for missed or delayed diagnosis. Furthermore, barriers to provision of face-to-face diagnostic and intervention services during the pandemic may also limit the identification of ASD and delivery of therapies for eligible children.
      • Eshraghi A.A.
      • Li C.
      • Alessandri M.
      • et al.
      COVID-19: overcoming the challenges faced by individuals with autism and their families.
      These reductions in screening and care may compound pre-existing racial, ethnic, and sociodemographic disparities in ASD identification and intervention.
      • Wiggins L.D.
      • Durkin M.
      • Esler A.
      • et al.
      Disparities in documented diagnoses of autism spectrum disorder based on demographic, individual, and service factors.
      ,
      • Baio J.
      • Wiggins L.
      • Christensen D.L.
      • et al.
      Prevalence of autism spectrum disorder among children aged 8 years - autism and developmental disabilities monitoring network, 11 sites, United States, 2014.
      Lower rates of screening among children during the COVID-19 pandemic warrant attention.

      Limitations

      We were unable to examine the data for children who did not present for care during the pandemic; therefore, our findings may overestimate screening-completion rates of the entire practice population and may have excluded the groups of children at the highest risk of being unscreened (eg, selection bias). Our findings also may not generalize to practices without integrated electronic screening, or to patients with differential access to or comfort with electronic devices for screening completion. We only examined completion of one ASD screener; as in our prior work, even more children may have missed the second recommended screening at 24 months, which is important for improving ASD detection.
      • Guthrie W.
      • Wallis K.
      • Bennett A.
      • et al.
      Accuracy of autism screening in a large pediatric network.
      Lastly, we did not examine longitudinal screening outcomes. Therefore, we cannot determine if patient care and screening were delayed beyond our study timeframe or missed entirely. Future work can examine site-level variation in screening to better understand the impact of site-specific processes.

      Conclusions

      The impact of the COVID-19 pandemic on children with ASD requires further study. We identified some decreases in well-child care, particularly for Black children, and decreased screening completion across most racial and ethnic groups. Strategies for catch-up may be needed to screen children who missed the traditional screening window. Screening tools, such as the M-CHAT-F, are validated up to 30 months. Practices may consider adding opportunities to screen up until that age if a child missed an earlier screen. Attention to the developmental needs of traditionally underserved populations, including minoritized racial and ethnic groups, non-English speakers, and publicly insured populations, remains critical to ameliorate persistent disparities.

      References

      1. Pediatrician life and career experience study (PLACES) Special Issue. In: AAP, ed. Going Places Newsletter June 2020.

        • Ciacchini B.
        • Tonioli F.
        • Marciano C.
        • et al.
        Reluctance to seek pediatric care during the COVID-19 pandemic and the risks of delayed diagnosis.
        Italian J Pediatr. 2020; 46: 87
        • Hyman S.L.
        • Levy S.E.
        • Myers S.M.
        Identification, evaluation, and management of children with autism spectrum disorder.
        Pediatrics. 2019; e20193447
        • Guthrie W.
        • Wallis K.
        • Bennett A.
        • et al.
        Accuracy of autism screening in a large pediatric network.
        Pediatrics. 2019; 144e20183963
        • Dennis-Heyward E.A.
        • Shah S.N.
        Pediatric COVID-19 disparities and prioritizing equity—children are not spared.
        JAMA Pediatr. 2021; 175: 898-900
        • Brown C.L.
        • Montez K.
        • Amati J.B.
        • et al.
        Impact of COVID-19 on pediatric primary care visits at four academic institutions in the Carolinas.
        Int J Environ Res Public Health. 2021; 18: 5734
        • Robins D.L.
        • Fein D.
        • Barton M.L.
        • et al.
        The Modified Checklist for Autism in Toddlers: an initial study investigating the early detection of autism and pervasive developmental disorders.
        J Autism Dev Disord. 2001; 31: 131-144
        • Robins D.L.
        • Casagrande K.
        • Barton M.
        • et al.
        Validation of the modified checklist for Autism in toddlers, revised with follow-up (M-CHAT-R/F).
        Pediatrics. 2014; 133: 37-45
        • Khowaja M.K.
        • Hazzard A.P.
        • Robins D.L.
        Sociodemographic barriers to early detection of autism: screening and evaluation using the M-CHAT, M-CHAT-R, and Follow-Up.
        J Autism and Dev Disord. 2015; 45: 1797-1808
        • Trent M.
        • Dooley D.G.
        • Dougé J.
        The impact of racism on child and adolescent health.
        Pediatrics. 2019; 144e20191765
        • Van Dyke M.E.
        • Mendoza M.C.B.
        • Li W.
        • et al.
        Racial and ethnic disparities in COVID-19 incidence by age, sex, and period among persons aged <25 years - 16 U.S. Jurisdictions, January 1-December 31, 2020.
        MMWR Morbid Mortal Wkly Rep. 2021; 70: 382-388
        • Craig S.
        • Shen A.K.
        • Wallis K.
        • et al.
        How health systems can help address language barriers to achieve digital health equity.
        PolicyLab Blog. 2021; (Available at: https://policylab.chop.edu/blog/how-health-systems-can-help-address-language-barriers-achieve-digital-health-equity. Accessed May 13, 2022)
        • Eshraghi A.A.
        • Li C.
        • Alessandri M.
        • et al.
        COVID-19: overcoming the challenges faced by individuals with autism and their families.
        Lancet Psychiatry. 2020; 7: 481-483
        • Wiggins L.D.
        • Durkin M.
        • Esler A.
        • et al.
        Disparities in documented diagnoses of autism spectrum disorder based on demographic, individual, and service factors.
        Autism Res. 2020; 13: 464-473
        • Baio J.
        • Wiggins L.
        • Christensen D.L.
        • et al.
        Prevalence of autism spectrum disorder among children aged 8 years - autism and developmental disabilities monitoring network, 11 sites, United States, 2014.
        MMWR Morbid Mortal Wkly Rep. 2018; 67 (Surveillance summaries (Washington, DC : 2002)): 1-23