Leveraging Novel Clinical Decision Support to Improve Preferred Language Documentation in a Neonatal Intensive Care Unit

Abstract

Background Recognition of the patient and family's diverse backgrounds and language preference is critical for communicating effectively. In our hospital's electronic health record, a patient or family's identified language for communication is documented in a discrete field known as “preferred language.” This field serves as an interdepartmental method to identify patients with a non-English preferred language, creating a bolded banner for non-English speakers easily identifiable by health care professionals. Despite the importance of documenting preferred language to facilitate equitable care, this field is often blank.

Objectives Using the Institute for Healthcare Improvement's Model for Improvement, our team sought to increase preferred language documentation within the neonatal intensive care unit (NICU) from a baseline of 74% in September 2021 to above 90% within 6 months.

Methods A multidisciplinary team was assembled to address preferred language documentation. Our team incorporated guidance regarding preferred language documentation into a novel clinical decision support (CDS) tool aimed at addressing various safety and quality measures within the NICU. Our primary outcome metric was documentation of family's preferred language. Process measures included overall compliance with the CDS tool. A secondary outcome was the assessment of preferred language documentation accuracy.

Results The average rate of preferred language documentation increased from a baseline of 74 to 92% within 6 months and is currently sustained at 96%. Moreover, our follow-up assessments found that 100% of a random sample of contacted patients (n = 88) had their language preferences documented correctly. Overall compliance with the CDS tool remained at 85% throughout the project.

Conclusion Using a quality improvement framework coupled with a CDS initiative, our team was able to successfully and accurately improve preferred language documentation in our NICU. Future projects include strategies for more equitable care for patients with non-English preferences such as improved interpreter usage and discharge instructions in their preferred language.

 

MissingLinkText

Keywords

clinical decision support - preferred language - health care systems electronic health record - neonatal

Background and Significance

As the United States population continues to grow and become more diverse, health care systems face challenges in optimizing medical care for a multilingual population.[1] In the most recent American Community Survey completed in 2021, 21.6% of individuals 5 years of age or older identified as speaking a language other than English (LOE) at home, with 38.3% of this subset of the population listing their ability to speak English less than “very well.”[2] Timely and accurate identification of patients with an LOE holds significant importance in delivering language-appropriate care and aligning principles of equity and fairness.[3] [4]

Multiple studies have demonstrated numerous health disparities in patients with a non-English preferred language for communication. These health disparities include more serious adverse events, longer hospital stays, more frequent emergency department visits, and decreased home health care referrals.[3] [4] [5] [6] [7] [8] [9] Language-appropriate services therefore become paramount as families who prefer an LOE are also at an increased risk of receiving less information about their child's care.[10] It has also been shown that the use of interpreters can range from 36 to 77% for all discussions with LOE families, indicating a substantial communication disparity.[11] [12] Limiting disparities through the provision of language-appropriate services such as language-concordant care (i.e., services provided by a clinician who speaks the same language) or interpreter-mediated care starts with a commitment to appropriately identifying LOE families.[5]

The U.S. Department of Health and Human Services mandates the collection of race, ethnicity, and language data from patients within the electronic health record (EHR).[13] Without appropriate and accurate documentation of language preference within the EHR, patients and their families are at an increased risk of the aforementioned disparities. This lack of documentation also hinders opportunities for quality improvement (QI) and research efforts aimed at mitigating these risks.[14] [15] [16] Previous studies have indicated that the accuracy of the EHR in identifying families with a non-English preference can vary significantly, with positive predictive values ranging from 60 to 100%.[14] After recognizing the incomplete state of language preference documentation in our EHR and unknown accuracy of the existing data, our group sought to improve our language documentation to better inform quality care initiatives and improve our delivery of care to an underrecognized population.

 

MissingLinkText

Objectives

Our multidisciplinary team's primary goal was to improve the documentation of preferred languages within the EHR, ensuring comprehensive coverage of both English and non-English preferences. Using the Institute for Healthcare Improvement's (IHI) Model for Improvement, a QI framework, we aimed to increase our baseline documentation from 74% in September 2021 to above 90% within a 6-month timeframe.[17] Secondarily, we assessed the accuracy of preferred language documentation.

 

MissingLinkText

Methods

Context

The Children's Hospital of Philadelphia neonatal intensive care unit (NICU) is a large, urban, academic facility caring for a wide variety of critically ill infants with disease processes including complications of extreme prematurity, congenital anomalies requiring surgical intervention, and referral for subspecialty care. Our NICU primarily serves infants referred from other hospitals or admitted through our emergency department, with fewer inborn patients. The NICU consists of a diverse patient population with over 85 spoken languages.[18] Within the Epic EHR (Verona, Wisconsin, United States), preferred language is a discrete data element that is stored at the patient level. Entry of preferred language into the designated field has downstream effects including updating the user interface notifying providers, nurses, and other allied health staff of non-English preferred patient languages ([Fig. 1]). Preferred language is documented for the primary caregiver of the infant and, in current state, does not account for multilingual households.

Fig. 1 Electronic health record screenshot demonstrating how updating the preferred language field within a patient's demographic form changes the patient's page to display language other than English preference for health professionals. 2023 Epic Systems Corporation. Data presented in the figure are imaginary.

 

MissingLinkText

Intervention

A multidisciplinary team comprised of NICU leadership, informatics experts, nursing, QI leaders, advanced practice providers, and NICU physicians assessed key drivers contributing to the deficiency of preferred language documentation ([Fig. 2]). Our multidisciplinary team explored change concepts which could be implemented to achieve our primary goal. Using the IHI Model, in our first plan-do-study-act (PDSA) cycle, our team incorporated guidance regarding preferred language documentation into a novel clinical decision support (CDS) tool aimed at addressing various safety and quality measures within the NICU. This tool consists of a set of daily safety and QI-focused questions that are embedded in the EHR, tailored to each patient, and are a part of the established workflow in our NICU. The CDS tool is completed by the NICU clinician (i.e., advanced practice practitioner, resident physician, fellow physician, or attending physician) at the bedside during morning rounds. The CDS tool is noninterruptive, but its design includes flags on patient lists indicating whether it has been completed for the day. The Epic-based version of this CDS tool has been active in our NICU since October 2019.[19] Additional PDSA cycles focused on data accuracy.

Fig. 2 Key driver diagram showing our primary goal and associated drivers with change concepts to achieve our goal.

In the first PDSA cycle, the question “Is the family's preferred language appropriately documented?” was added to the CDS tool. Within the tool itself, we included instructions and embedded a hyperlink to guide clinicians directly to the preferred language field for ease of completion ([Fig. 3]). To lessen the workload burden, we built logic within the tool to ensure that once the preferred language field was completed it would not be asked again. The question would repeat itself weekly if the preferred language field was not completed. Providers and NICU staff were made aware of the upcoming changes to the CDS tool before the intervention went live on September 30, 2021.

Fig. 3 Example of a weekly question utilizing our CDS tool in the electronic health record. The preferred language question is outlined which was introduced during PDSA Cycle 1.

NICU staff were updated on overall compliance with the CDS tool during monthly, unit-wide conferences. Progress related to this preferred language project was specifically discussed following the second PDSA cycle.

 

MissingLinkText

Measures and Statistical Analysis

Our primary outcome was the percentage of patients in our unit with complete preferred language documentation. This was analyzed in weekly increments and plotted on a statistical process control (SPC) P-chart within Microsoft Excel (2023). As per QI standards, special cause variation was determined to be present if there were eight consecutive points above or below the centerline, six consecutive points increasing or decreasing, fifteen points within one standard deviation of the centerline, a point noted outside the control limits, or two out of three consecutive points noted near the control limits.[17]

As a secondary aim, the accuracy of preferred language documentation was assessed using direct discussion with patient families. At approximately 6-month intervals from CDS implementation, our team called 30% of randomly chosen primary caregivers/guardians of admitted patients (n = 92). Division QI experts determined this 30% sample. Duplicate families from previous follow-ups were excluded, and new families were contacted. If a LOE was noted in the EHR, a hospital-based phone interpretation service assisted with calling the family. Follow-up data were collected and stored utilizing a REDCap database.[20] [21] With these subsequent PDSA cycles, caregivers/guardians were asked three questions: (1) whether the preferred language documented in the EHR was accurate, (2) if the NICU staff communicates with them in their preferred language, and (3) if they remembered being asked about their preferred language. These responses were analyzed using descriptive statistics.

Our primary process metric was overall compliance with the CDS tool, which was tracked and plotted in monthly increments on an SPC P-chart. Annual revisions, or CDS tool updates, were completed with input from NICU wide-surveys and leadership to determine if questions that should be added, revised, or removed to align with current NICU goals. Initial data collection on compliance for the CDS tool were obtained over weeks and transitioned to months when appropriate.

 

MissingLinkText

 

MissingLinkText

Results

In coordination with our intervention, preferred language documentation increased from a baseline of 74 to 92% within 6 months. Moreover, this improvement in documentation has been sustained and is currently at 96%, as depicted on the P-chart for preferred language documentation ([Fig. 4]). During this time, the preferred language question in our CDS tool had a completion rate of 89.3%. Overall compliance with the CDS tool remained at 85% throughout the project which is depicted on our P chart for CDS tool compliance ([Fig. 5]).

Fig. 4 P-chart showing overall improvement in the primary outcome metric of preferred language documentation since the introduction of our intervention question. The current centerline (mean) is 96%, improved from a baseline of 74%.

Fig. 5 P-chart showing process measure of overall compliance with the CDS tool within the NICU since development. The initial baseline is short due to our transition from weekly to monthly evaluations as the project has extended over multiple years. The current centerline (mean) is 85%, improved from a baseline of 31%.

Across three follow-up assessments, a total of 92 primary caregivers/guardians were called (n = 31 at 6-month assessment, n = 31 at 12-month assessment, and n = 30 at 18-month assessment). Of those 92 called, four were unable to be reached via telephone, leaving 88 families in our analysis. English was the predominant preferred language (87.5%), with Spanish (6.8%), Haitian Creole (1.1%), Uzbek (3.4%), and Nepali (1.1%) also being represented ([Table 1]). Preferred language was correctly documented 100% of the time based on verbal confirmation with the primary caregiver/guardian. Moreover, the NICU was reported to be communicating in the preferred language 100% of the time. When all caregivers/guardians were asked if they recalled being asked about their preferred language, 30.6% said yes (n = 27), 33.0% said no (n = 29), and 36.4% were unsure (n = 32, [Table 2]). When focusing on responses from caregivers/guardians with non-English preferences, 82% (9 out of 11) remembered being asked about their preferred language.

Table 1 Breakdown of preferred language of primary caregiver/guardian in preferred language follow-up assessments

Preferred language of primary caregiver/guardian ( N  = 88)	N (%)
English	77 (87.5%)
Spanish	6 (6.8%)
Uzbek	3 (3.4%)
Haitian Creole	1 (1.1%)
Nepali	1 (1.1%)

Note: Pooled results across the three different assessments, total n = 88, approximately 30% of patients admitted to the neonatal intensive care unit called for each follow-up assessment (92 caregivers/guardians called and 4 were unable to be reached).

Table 2 Responses of primary caregiver/guardian to questions regarding preferred language documentation and NICU communication

Response ( N  = 88)	N (%)
It says in our electronic medical record that your preferred language for medical updates is ________, is this correct for the primary caregivers of your son/daughter?
Yes	88 (100%)
No	0 (0%)
Is the NICU currently communicating with you in your preferred language?
Yes	88 (100%)
No	0 (0%)
Did someone from the NICU ask you about your preferred language?
Yes	27 (30.6%)
No	29 (33.0%)
Unsure	32 (36.4%)

Abbreviation: NICU, neonatal intensive care unit.

Note: Pooled results across the three different assessments, total n = 88, approximately 30% of patients admitted to the NICU called for each follow-up assessment.

 

MissingLinkText

Discussion

With the incorporation of a language-focused question into our EHR-based QI/patient safety rounding checklist, preferred language documentation improved within 6 months. Moreover, this improvement has been sustained since the introduction of this question. Through follow-up interviews, the preferred language documentation field was noted to be 100% accurate for a total of 88 families/caregivers across three separate time points. We felt it was important to not only improve the amount of documentation, but also assess the accuracy of this process, given the awareness that the EHR documentation of preferred language does not guarantee accuracy.[14]

Our QI work highlights the effectiveness of utilizing established CDS to assist in new QI efforts. With small CDS changes, we were able to guide the NICU staff toward our goal of improved preferred language documentation, without explicitly stating our goal initially. The project's success is attributed to effectively integrating a preestablished CDS intervention into the existing routines of our unit's health care providers for new QI measures. Implementation of CDS that does not align with a provider's workflow can cause alert fatigue or decreased effectiveness.[22] Our intervention aligns with the “five rights of CDS” as it conveys the appropriate workflow for preferred language documentation (right information), to the health care provider (right person), via safety and QI-focused questions embedded in EHR (right format), using existing CDS (right channel) during rounds (right time).[23]

Interestingly, there was a relatively uniform distribution of caregivers/guardians that reported that they recalled, did not recall, or were unsure about being asked about their preferred language. Many families who were unsure attributed this to the lag since the time of admission or to competing factors during the admission process. However, our finding that only 82% of caregivers/guardians with non-English preference remembered being asked about their language preference suggests that these families may not be approached regarding their preferred language during a hospitalization. Additionally, for English-speaking families, only 30% recalled being asked about their preferred language. Assuming an English preference, without confirmation, may overlook a family's preference for a LOE, undermining health care equity. Appropriately identifying patients with non-English preference is vital to being able to provide equitable services, and it is concerning that a proportion of this population may not have discussed language preference with their care team. However, it is possible that they may not have been able to remember this component of their hospital-based care. The completion of the language preference field should require confirmation with a family, even if an interpreter is being used, and should not be biased based on interactions with the family. In future EHR states, incorporating self-reported language fields into patient intakes could potentially alleviate the problem described above and help better identify non-English language preferences. This establishes a systematic check to confirm language preferences and empowers families, particularly bilingual ones, to speak up if they are misidentified as English-preferred.

Ultimately, with increased, accurate preferred language documentation, we now can use this information to guide future efforts to improve language-informed care. With enhanced documentation and ongoing data collection, we can further elucidate the impact of preferred language on patient/family interaction with the health care system. For example, this may include efforts focused on the family utilization of online patient portals, where there are known disparities based on patient-preferred language.[24] Looking ahead, our group hopes to complete audits within our NICU to assess the appropriate use of translators and the distribution of translated documentation.

Limitations

This project was limited by several factors including the structure of the preferred language field. Due to the parent–child dyad in pediatrics, one limitation of the preferred language field is its inability to differentiate between the language preference of individual caregivers and guardians. Moreover, we did not distinguish between spoken and written language preferences for our NICU families. Our follow-up assessment was limited by the utilization of the primary phone number listed in the chart, which does not distinguish between parental figures (e.g., mother, father, grandparent, guardian, etc.) as the primary caregiver. In addition, we are unable to comment on any baseline accuracy of language preference documentation. Our initial accuracy assessment was completed following our CDS tool implementation, so we are unable to comment on whether accuracy was improved following this intervention.

In addition, our project likely suffers from limited generalizability. Our group recognizes that this project's success was largely due to the prior implementation of CDS infrastructure that is now part of our NICU's rounding culture. During the initial deployment of the CDS tool from August 2018 through March 2020, unannounced observers were employed to mitigate the Hawthorne effect.[19] However, with the initiation of our project in October 2021, providers were receiving regular feedback about NICU QI initiatives related to this CDS tool at unit-wide conferences. We anticipate that providers are aware that they are being observed, which may impact their behavior. With an overall compliance rate of 85%, there is a notable commitment by our NICU staff to complete the quality and safety checklist. Lastly, we noted that the population cared for in our unit is largely English-speaking and located at an academic urban center, not representative of all NICU populations in the United States.

 

MissingLinkText

 

MissingLinkText

Conclusion

Using a QI framework coupled with a novel CDS initiative, our team was able to successfully improve preferred language documentation in our NICU. This documentation was noted to be accurate upon confirmation with the patient's caregiver/guardian. Initial work such as this lays the foundation for future projects aiming to provide more equitable care for patients with non-English preferences such as improved interpreter usage and discharge instructions in their preferred language.

 

MissingLinkText

Clinical Relevance Statement

Knowing the language preferences of one's patient and their family is crucial in being able to provide effective and equitable clinical care, especially when non-English languages are involved. These patients face risks of worse health outcomes and accurate identification of family's preferred language is important to foster a more equitable health care setting. By adopting a QI-centered mindset and leveraging a CDS tool, our group was able to successfully and accurately increase the preferred language documentation in our NICU.

 

MissingLinkText

Multiple-Choice Questions

1. What is the definition of “language concordant care” in health care?

   • (a) a type of care that focuses on concordance between the ethnicity and race of the health care provider and patient

   • (b) care provided by health care professionals who speak multiple languages fluently

   • (c) a culturally sensitive approach where health care providers and patients speak the same language, allowing for effective communication and understanding

   • (d) A care model that emphasizes the use of interpreters and translators in all aspects of a patient's medical care when necessary.

   Correct Answer: The correct answer is option c. Language concordant care refers to a culturally sensitive approach where health care providers and patients speak the same language, allowing for effective communication and understanding. This approach is crucial in providing high-quality care to patients with a preferred language other than English. By eliminating language barriers, patients can better express their symptoms, concerns, and medical histories, leading to improved diagnoses, treatment plans, and overall health outcomes.

2. What data does the U.S. Department of Health and Human Services mandate collection of?

   • (a) race

   • (b) ethnicity

   • (c) language preference

   • (d) all of the above

   Correct Answer: The correct answer is option d. The U.S. Department of Health and Human Services mandates the collection of race, ethnicity, and language data from patients within the electronic health record. The collection of such data is the first step in being able to start to mitigate health disparities.

 

MissingLinkText

 

MissingLinkText

Conflict of Interest

None declared.

Acknowledgments

O.M. and A.R. sincerely acknowledge our mentors for propelling us toward greater things, and our patients who inspire us to aim for more equitable care.

Protection of Human and Animal Subjects

The study was conducted under local institutional standards for quality improvement initiatives and therefore institutional review boards approval was not needed.

• References

• 1 Bharmal A, Earl N, Klaber B. Improving equity in inpatient paediatric services: a single-centre quality improvement project providing care in patients' preferred language. BMJ Paediatr Open 2022; 6 (01) e001525
• 2 U.S. Census Bureau American Community Survey: Detailed languages spoken at home. Accessed July 20, 2023 at: https://data.census.gov/table?q=language&tid=ACSST1Y2021.S1601
• 3 Diamond L, Izquierdo K, Canfield D, Matsoukas K, Gany F. A systematic review of the impact of patient-physician non-English language concordance on quality of care and outcomes. J Gen Intern Med 2019; 34 (08) 1591-1606
• 4 Yeboah D, McDaniel C, Lion KC. Language matters: why we should reconsider the term “limited English proficiency”. Hosp Pediatr 2023; 13 (01) e11-e13
• 5 Ortega P, Shin TM, Martínez GA. Rethinking the term “limited English proficiency” to improve language-appropriate healthcare for all. J Immigr Minor Health 2022; 24 (03) 799-805
• 6 Hartford EA, Anderson AP, Klein EJ, Caglar D, Carlin K, Lion KC. The use and impact of professional interpretation in a pediatric emergency department. Acad Pediatr 2019; 19 (08) 956-962
• 7 Cohen AL, Rivara F, Marcuse EK, McPhillips H, Davis R. Are language barriers associated with serious medical events in hospitalized pediatric patients?. Pediatrics 2005; 116 (03) 575-579
• 8 Ngai KM, Grudzen CR, Lee R, Tong VY, Richardson LD, Fernandez A. The association between limited English proficiency and unplanned emergency department revisit within 72 hours. Ann Emerg Med 2016; 68 (02) 213-221
• 9 Levas MN, Cowden JD, Dowd MD. Effects of the limited English proficiency of parents on hospital length of stay and home health care referral for their home health care-eligible children with infections. Arch Pediatr Adolesc Med 2011; 165 (09) 831-836
• 10 Thornton JD, Pham K, Engelberg RA, Jackson JC, Curtis JR. Families with limited English proficiency receive less information and support in interpreted intensive care unit family conferences. Crit Care Med 2009; 37 (01) 89-95
• 11 Lion KC, Gritton J, Scannell J. et al. Patterns and predictors of professional interpreter use in the pediatric emergency department. Pediatrics 2021; 147 (02) e20193312
• 12 Gupta KM, Campeggio D, Madu C. et al. Improving identification of interpreter need in the pediatric emergency department. Pediatrics 2023; 151 (03) e2022057330
• 13 Centers for Medicare and Medicaid Services. Inventory of Resources for Standardized Demographic and Language Data Collection. Accessed July 20, 2023 at: https://www.cms.gov/About-CMS/Agency-Information/OMH/Downloads/Data-Collection-Resources.pdf
• 14 Rajaram A, Thomas D, Sallam F, Verma AA, Rawal S. Accuracy of the preferred language field in the electronic health records of two Canadian hospitals. Appl Clin Inform 2020; 11 (04) 644-649
• 15 Klinger EV, Carlini SV, Gonzalez I. et al. Accuracy of race, ethnicity, and language preference in an electronic health record. J Gen Intern Med 2015; 30 (06) 719-723
• 16 Chin MH. Using patient race, ethnicity, and language data to achieve health equity. J Gen Intern Med 2015; 30 (06) 703-705
• 17 Ogrinc GS, Headrick LA, Barton AJ. et al. Fundamentals of Health Care Improvement: A Guide to Improving Your Patients' Care (4th edition). Joint Commission Resources and Institute for Healthcare Improvement;; 2022
• 18 Office of Innovation and Technology. City of Philadelphia: PHL Language Services Usage Dashboard. Accessed July 20, 2023 at: https://experience.arcgis.com/experience/51923768a6b14d5bb773bfcc1cf74cd1/?mc_cid=d55b2adff2&mc_eid=2fafe475a7
• 19 Carr LH, Padula M, Chuo J. et al. Improving compliance with a rounding checklist through low- and high-technology interventions: a quality improvement initiative. Pediatr Qual Saf 2021; 6 (04) e437
• 20 Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42 (02) 377-381
• 21 Harris PA, Taylor R, Minor BL. et al; REDCap Consortium. The REDCap consortium: building an international community of software platform partners. J Biomed Inform 2019; 95: 103208
• 22 Chaparro JD, Beus JM, Dziorny AC. et al. Clinical decision support stewardship: best practices and techniques to monitor and improve interruptive alerts. Appl Clin Inform 2022; 13 (03) 560-568
• 23 Campbell R. The five “rights” of clinical decision support. J AHIMA 2013; 84 (10) 42-47 , quiz 48
• 24 Sun EY, Alvarez C, Callahan LF, Sheikh SZ. The disparities in patient portal use among patients with rheumatic and musculoskeletal diseases: retrospective cross-sectional study. J Med Internet Res 2022; 24 (08) e38802

Publikationsverlauf

Eingereicht: 05. Oktober 2023

Angenommen: 22. Mai 2024

Accepted Manuscript online:
24. Mai 2024

Artikel online veröffentlicht:
31. Juli 2024