Communication Practices for Families With Languages Other Than English in US Neonatal Care Units

We surveyed clinical leaders from level 2 to 4 NICUs. Using data from the 2018 American Hospital Association list of US hospitals, we randomly selected 100 NICUs across 5 US regions (n = 500) to ensure the representation of practices across different settings. A full description of sample size estimation and attainment of the final hospital sample are detailed elsewhere.¹¹  The survey inquired about hospital attributes, systematic data collection regarding families’ language information, usage patterns of language services, and barriers to the provision of CLAS. Between January and November 2022, we distributed the ∼15-minute REDCap survey via an e-mail link.

Our main outcomes included consistent parental language documentation (≥75% of the time versus <75%) and consistent professional interpreter use (in-person or remote interpretation ≥75% of the time versus <75%). Potential predictors included geographic region (Northeast, Midwest, Southeast, Southwest, West), hospital characteristics (hospital type, teaching status, safety-net status, rural versus urban), NICU characteristics (highest level of care, annual admissions), and patient mix (proportion of NICU admissions that were Hispanic/Latinx or Black or that had LOE).

To obtain national prevalence estimates, analyses were adjusted by using weights to factor in the probability of hospital selection from the entire eligible set of hospitals within each region and to account for variations in response rates across regions. We used χ² tests and bivariate logistic regression to examine the associations of potential predictors and our main outcomes. We set the level of significance at α = 0.05 and present results using odds ratios and 95% confidence intervals. All analyses were conducted by using SAS procedures for complex survey designs in SAS 9.4 (SAS Institute, Cary, North Carolina, USA). This study was approved by our university’s institutional review board.

The overall response rate was 34%. The characteristics of participating hospitals are presented in Table 1. For our main outcomes, 63% of NICUs consistently documented language information, whereas only 41% of NICUs reported using any modality of interpretation to communicate with families with LOE ≥75% of the time (Table 2). There was significant variability by service hours in the frequency of use of various interpretation modalities (Fig 1). During regular hours, in-person interpreters were underutilized compared with telephonic or video services. Approximately half of NICUs reported relying on ad hoc interpretation “sometimes” during regular and off hours. Another source of practice variation was the type of communication scenario, with professional in-person interpreters more frequently employed during family conferences and discharge teaching, yet notably underused during daily clinical rounds and regular bedside updates (Table 3).

Analyses of potential predictors of our main outcomes revealed that teaching (versus nonteaching) hospital status, highest level of neonatal care, and larger NICU size were associated with consistent language documentation (Supplemental Table 4). Only a larger NICU size was associated with consistent interpreter use (Supplemental Table 5). Notably, the proportion of patients with LOE did not significantly influence the likelihood of language data documentation or interpreter service utilization.

Surveyed clinical leaders reported barriers to the provision of CLAS at multiple levels. At the provider level, barriers included a lack of knowledge on how to access language services and challenges related to timely access to services in the context of time pressures related to the NICU environment. Team- and family-level barriers included few bilingual clinical providers and lower caregiver presence in the NICU among families speaking LOE compared with English-speaking families. Among system-level barriers, respondents indicated a lack of appropriate language services during off hours and for uncommonly spoken languages and poor quality of certain modalities of interpretation.

Despite longstanding federal mandates and the clear benefits associated with professional interpretation and language-concordant care, we found a largely suboptimal and highly variable provision of CLAS for families with LOE in a nationally representative sample of US NICUs. These findings highlight a systemic evidence-practice gap that not only hinders care safety and quality but also perpetuates unequal treatment based on language needs in the NICU setting.

We found inconsistent documentation of language information from non-English-speaking families, along with variability in the language measures collected. A shift toward nondeficit-oriented terminology such as “language other than English (LOE)” or “non-English language preference (NELP)” has been proposed to reduce stigma and promote patient-centeredness.^1,12  Overall, there is a need for standardized measures and procedures to ensure that language data collection is thorough, accurate, and used effectively.¹³  Practices such as the verification of language information from both caregivers on admission to the NICU and language identification cards displayed before entering patients’ rooms have the potential to increase awareness of families’ language needs and prompt appropriate use of interpreter services.¹⁴ 

Similar to reports in other pediatric care settings,¹⁵  we found underutilization of professional interpretation and continued reliance on ad hoc interpreters,¹⁶  reflecting systemic barriers that extend beyond individual NICUs. The fact that high-quality interpretation is most frequently used for family conferences and discharge planning but significantly less so for daily clinical updates or clinical rounds indicates a disconnection between the recognized importance of these services and their actual application in routine care.¹⁷  In addition, the differential use of interpretation services based on the time of day and type of communication event highlights the importance of language support services that are flexible and responsive to families’ changing needs throughout the NICU hospitalization.^18,19 

Our findings that larger hospital size and highest level of NICU care were associated with adherence to CLAS standards may reflect better-resourced settings’ ability to enforce mandates. That teaching hospital status was associated with consistent language documentation may reflect academic organizations’ commitment to a culture of improvement.²⁰  The operationalization of CLAS standards in lower-resourced hospital settings that often serve a disproportionate number of families with LOE remains difficult and requires particular attention.²¹ 

Strengths of this study include the sampling strategy and weighting procedures that enabled the estimate of the national prevalence of communication care practices. Limitations include social desirability and recall bias inherent to surveys. Notwithstanding, our findings contribute to understanding the current patterns of provision of language services in NICUs to inform future efforts to reduce linguistic disparities and improve the quality of care for a growing multilingual population.

Addressing barriers to CLAS provision requires a multifaceted approach. In addition to organizational commitment and investment to enforce CLAS mandates, recommendations include developing interventions tailored to the NICU context that tactically employ different modalities of interpretation according to the communication needs of families throughout hospitalization, workforce diversification, and ensuring that NICU staff are trained and equipped to address language barriers effectively. To move forward, NICUs must prioritize language justice as part of their commitment to high-quality, equitable care.

We wish to acknowledge Emma Jacobs, Selena Chen, Andrea Kuriyama, Farah Delgado, and Mariana Serrano for their critical assistance in obtaining clinical leader’s contact information that enabled electronic survey administration.