Telemedicine Improves Rate of Successful First Visit to NICU Follow-up Clinic Free

We performed a retrospective chart review to ascertain baseline rates of successful first NICU follow-up clinic visit by 12 months’ corrected age. The early cohort was born between January 2018 and June 2019 and the late cohort was born between May 2020 and May 2021. Telemedicine consults were available for infants in the late cohort.

The Barbara Bush Children’s Hospital at Maine Medical Center has a 51-bed level III NICU in a city with a population of 68 400. The goal of the NICU follow-up clinic is to provide routine developmental surveillance to the highest risk infants.

In the early cohort, nurse care managers would approach parents about the opportunity for attending the outpatient NICU follow-up clinic as the infant neared discharge. There was no formal process in place for parents who declined participation. For parents who agreed to attend the NICU follow-up clinic, documentation in discharge paperwork was not standardized. The discharge summary might include a recommendation for attending the NICU follow-up clinic but would not have provided contact information for the clinic or an appointment date and time. There were no direct interactions between the developmental and behavioral pediatrician and the parents before NICU discharge. This was primarily because outpatient clinic responsibilities did not permit the developmental and behavioral pediatrician to make rounds in the NICU.

In response to the COVID-19 pandemic, by the end of April 2020, it was possible for the late cohort to receive reimbursable telemedicine consultation with the developmental and behavioral pediatrician before discharge from the NICU. Coordinating this inpatient telemedicine consult required additional steps. Parents were approached by a nurse care manager in the weeks before anticipated discharge to ascertain interest and availability for the telemedicine consult. This information was communicated to the developmental and behavioral pediatrician and to the NICU team. The NICU team placed the inpatient consult order, and the developmental and behavioral pediatrician scheduled the telemedicine meeting with the family. On completion of the inpatient telemedicine consult, the developmental and behavioral pediatrician would place an appointment reminder for outpatient NICU follow-up in the electronic medical record (EMR). Documentation of the appointment reminder in the EMR ensured the appointment would be communicated to the primary care provider (PCP) in the discharge summary and to the parents in the after-visit summary (AVS). In the event of parental refusal for NICU follow-up referral, the neonatologist on service would be notified and would then discuss the importance of NICU follow-up with the parents.

Infants included were those who survived NICU admission and met the NICU follow-up clinic referral criteria:, gestational age ≤29 6/7 weeks, a birth weight <1500 g, and grade III/IV intraventricular hemorrhage, stroke, seizure, hypoxic ischemic encephalopathy, or nitric oxide therapy for severe pulmonary hypertension.

Infants who died before discharge from the NICU, who relocated after discharge, or who died during the first 12 months of life were excluded.

The developmental and behavioral pediatrician met with families via Zoom. Verbal consent was obtained from the parent for the telemedicine consult and patient identification was confirmed via date of birth. The developmental and behavioral pediatrician spent an average of 20 minutes with families to discuss the role of the NICU follow-up clinic, the importance of monitoring the child’s neurodevelopment after discharge, how early interventions can improve outcomes, and available support for children with special health care needs.

Demographic and hospitalization information was acquired from the EMR. The rurality of each participant’s primary residence was determined via the US Department of Agriculture’s Rural-Urban Commuting Area codes, which uses measures of population density, urbanization, and daily commuting to map ZIP codes to urban or rural categories. We condensed the Rural-Urban Commuting Area classifications into 3 categories: metro, large rural, and small/isolated rural. The late cohort, which had telemedicine consults available before discharge, was compared with the early cohort using the Pearson χ² test, Fisher exact test, or Wilcoxon rank-sum test as appropriate. We used logistic regression to compare follow-up rates between cohorts, controlling for maternal insurance as a proxy for socioeconomic status. We also performed a mediation analysis to assess the relative contribution of NICU follow-up reminders for parents and PCPs to the late cohort’s improved follow-up rate. This study was approved under expedited review as a retrospective chart review by the Maine Medical Center institutional review board. All analyses were performed in R 3.6.2.

There were 364 infants identified by the EMR query (Fig 1, A and B). Of these, there were ultimately 152 in the early cohort and 105 in the late cohort who met inclusion criteria for the study. Mothers from the late cohort were slightly older and had a significantly higher rate of private insurance compared with mothers in the early cohort (Table 1). There were no significant differences in terms of primary maternal language, adherence with prenatal care, and either rurality or geographic distance from the hospital (Table 1). Maternal education level was not recorded in the EMR for 120 (79%) of the early cohort or for 85 (81%) of the late cohort. For those with documentation of educational level, there were no significant differences. The sex and gestational ages of the infants were not different between groups. For the late cohort, there was a significantly higher rate of referral to NICU follow-up documented in the AVS for parents (89% vs 60%, P < .001) and in the discharge summary for the PCP (91% vs 28%, P < .001).

There were 90 (86%) infants in the late cohort who received telemedicine consults before discharge. Reasons for not having telemedicine consults in the late cohort included back-transfer to a different hospital, earlier than expected discharge, or parental refusal. There were 40 (26%) children in the early cohort who attended the NICU follow-up clinic by adjusted age 12 months versus 64 (61%) in the late cohort (P < .001) (Table 1). After controlling for maternal insurance, the odds of attending a 12-month follow-up visit were 3.7 times higher for babies who had the availability of a telemedicine consult in the hospital (confidence interval [CI], 1.8-7.9). Approximately 37% of this effect was mediated by including information for referral for NICU follow-up in the AVS documentation for parents (P < .001). Documented referral for follow-up to the PCP did not mediate the relationship between cohort and follow-up (P = .31). After accounting for mediation by the change in documentation, there was still a significant effect of telemedicine on rate of follow up (odds ratio, 1.3; CI, 1.1-1.5).

This study took advantage of a natural experiment that occurred with the change in reimbursement for telemedicine visits as a result of the COVID-19 pandemic. We collected baseline data on the rate of NICU follow-up clinic attendance for a cohort of neonates (early cohort) who did not receive telemedicine consultation or standardized discharge recommendations for NICU follow-up clinic before hospital discharge. We then compared this cohort with a second cohort (late cohort) of neonates who not only had the opportunity to have a telemedicine consultation before discharge but also benefitted from an initiative to improve rates of follow-up through better documentation to parents and PCPs. We found that the availability of telemedicine consults substantially increased the rate of attendance at NICU follow-up clinic. We also showed that the improved documentation to parents, not PCPs, increased the likelihood of successful follow-up.

Challenges optimizing NICU follow-up clinic attendance are frequently reported and studies have identified many types of barriers to NICU follow up clinic attendance.¹¹  A qualitative research study identified lack of understanding of the importance of NICU follow-up and the fear of receiving more bad news as factors that seemed to play a role in the decision not to attend a NICU follow-up clinic.¹²  This study suggests that a single point of contact for families of NICU graduates, one who works across the locations of care (NICU, clinic, and even the community) could be considered as a way to overcome barriers. One of the advantages of the telemedicine intervention is that it does just that. It permits a single person, the outpatient developmental and behavioral pediatrician in our case, to work across multiple locations so the family is seeing the same familiar face. We speculate that families were able to build a more trusting relationship with the developmental and behavioral pediatrician as a result of meeting this person before hospital discharge. Although we did not gather data to support this, we believe that meeting through telemedicine may have also improved the parents’ education about the need for NICU follow-up. We are hopeful that insurance reimbursement for this type of telemedicine visit endures beyond the COVID-19 pandemic. Developmental and behavioral pediatricians are a limited resource outside of major academic centers and their sphere of influence can be substantially expanded when telemedicine can be used.

Other studies have found additional barriers to NICU follow-up clinic attendance relating to the social determinants of health that may be associated with racial groups, young maternal age, low income level, lower parental educational level, and greater traveling distance to follow-up clinics.^9,10  In our study, there were no significant differences in the maternal age between groups, and as Maine is a relatively homogenous state, we did not further investigate potential impacts of race and ethnicity as social determinants of health.¹³  Unfortunately, our institution does a poor job of capturing data on the education level of the mother, with essentially 80% missing data for both cohorts. This is disappointing given the strong data that associate maternal education level with infant outcome.^14,15  We did capture maternal health insurance data as a proxy for socioeconomic status and did find that the characteristics of the late cohort were different from the early cohort in this respect. More mothers in the late cohort had private rather than public health insurance, and we attribute this to changing regional demographics in our state as people of reproductive age migrated to Maine away from large metropolitan areas as a result of the pandemic. Interestingly, our results do not support that distance to the NICU follow-up clinic affects the rate of successful follow-up, which is different from other studies in California and Canada, where this has been a factor.^9,10 

Other interventions to improve the rate of NICU follow-up clinic attendance have been investigated. One quality improvement study focused efforts on educating NICU providers and parents about the importance of a NICU follow-up clinic appointment as well as improving care coordination and found improved attendance rates from 60% to 76%.¹⁶  A previous study has shown that implementing a multistep approach for discharge from the NICU, including parental education, increases likelihood of successful follow-up.¹³  Our study reinforces this finding: there was a significant difference in rate of successful follow-up after parents were educated by the developmental and behavioral pediatrician while still in the NICU and when referral was documented in the AVS to parents.

A strength of this study is the successful implementation of telemedicine consults for 86% of the infants in the late cohort as well as improved documentation for the need for NICU follow-up in the AVS for parents 89% of the time and in the discharge summary for PCPs 91% of the time. A weakness, however, is that we did not collect data on the reasons why telemedicine consults were not performed. Another weakness is that the maternal education level is not a variable routinely collected in our EMR. Our use of insurance status as a proxy for socioeconomic status is reasonable but has limitations. Last, the early cohort had a particularly low rate of follow-up for reasons that we attribute to lack of parental education about the importance of NICU follow-up and lack of contact with the developmental and behavioral pediatrician before hospital discharge. Unfortunately, we do not have the ability to know for certain that these are the reasons families did not attend a NICU follow-up clinic in the past. Future work could investigate the impact of improved NICU follow-up clinic attendance and the subsequent use of early intervention programs and improved school-age outcomes.

We acknowledge the contribution of Dr Samantha Dunn, the developmental and behavioral pediatrician, who completed all of the telemedicine consults.