Interdisciplinary Quality Improvement Project Increases Vitamin D Supplementation in Infants

Our project was completed at the Barbara Bush Children’s Hospital at Maine Medical Center, an embedded, urban, academic children’s hospital within a 600-bed tertiary care hospital in Portland, Maine. Our Family Birth Center includes a 17-bed Labor and Delivery unit, a 51-bed Level IV NICU/continuing care nursery, and a newborn nursery that accommodates 20 mother–infant dyads at a time. Each year 2300 to 2400 infants are discharged from the nursery, 30% of whom are cared for by the Pediatric Hospital Medicine Newborn Teaching Service (PHM-NTS). The remainder of the infants are cared for by a combination of a Family Practice Teaching Service and local private practices. Although proportionally small, 21% of PHM-NTS patients follow up in our onsite resident primary care clinic. This represents the largest percentage of infants discharged from the PHM-NTS following up in a single practice, and, therefore, we chose this practice as a convenience sample for evaluation of our outcome measure. Targeting patients of the pediatric resident clinic aided in the engagement and education of residents in the QI process. This resident clinic population is 40% English-speaking and 82% insured by Medicaid. Our institution has a comprehensive Interpreter Services department that provides live, telephone, or tablet device interpretation of discharge education for all non-English-speaking families.

In December 2016, an interprofessional improvement team began meeting monthly with a goal of increasing infant vitamin D supplementation rates in our pediatric resident clinic. The team included PHM physicians, pediatric residents, a nurse manager, a nurse educator, an inpatient pediatric pharmacist, and a community pharmacist.

Our interventions are outlined in the key driver diagram (Fig 1). We began with electronic health record (EHR) optimization (June 2017) and staff and family education (May 2017). Our outpatient community pharmacy, physically located within our hospital, offered bedside delivery of discharge medications and teaching by pharmacists. This bedside delivery service was not previously used in the nursery. To begin offering this service for vitamin D, we created an order set, including a prescription for oral liquid infant vitamin D and a community pharmacy consultation order for bedside teaching at the time of delivery. Free vitamin D was provided to families unable to pay out of pocket at the time of discharge (June 2016). To ensure proper routing of electronic prescriptions, our onsite community pharmacy was designated every infant’s “home pharmacy” at the time of birth. We used a best practice advisory (BPA), which was activated the first time a provider entered the chart of a newborn, to remind clinicians and link them to the order set (June 2017). During our monitoring, optimization, and maintenance period, we tracked process measure data for all nursery infants (Fig 2), and after we had achieved nursery-wide process measure levels >80%, we approached the nursery medical director and obtained permission to integrate the order set into standing protocol orders at time of birth (January 2020).

To address nursery staff awareness of the outpatient vitamin D recommendations, we provided in-service training at staff meetings using verbal and visual methods of education. Once the order set was in place, our unit-based nurse educator completed daily chart audits and provided targeted feedback to specific nurses. The pediatric resident on our team provided periodic feedback and updates to resident colleagues at regular residency meetings. In addition to the education provided to families at the bedside by pharmacists, the QI team modified an existing handout for parents about the importance of vitamin D supplementation in the first year of life (Supplemental Information). Similar information was included in the newborn written discharge instructions and reviewed with families by the discharging physician team. On the day of discharge, nurses assisted families with the administration of the first dose of vitamin D from their recently delivered “home” supply to model proper technique and provide needed feedback.

Additional interventions included creating systems for tracking the number of monthly discharges from the nursery and the number of families who received discharge medication delivery and teaching (January 2017). We created an outpatient EHR template capturing the details of the infant’s diet and medication history at well child visits, to discern whether parents reported adhering properly to supplementation recommendations (January 2017).

Interventions to cultivate continued engagement in this project included sharing process and outcome measure data by posting results in the staff lounge and discussing with residents at regular meetings. Interim results were presented at local QI and research forums and national nursing, pharmacy, and medical conferences.

Interventions were available nursery-wide and not restricted to the PHM-NTS patient population. Process measure was tracked for all patients, and the outcome measure was tracked in a smaller convenience sample of patients with follow-up in the resident clinic.

Our process measure numerator was the number of newborns from the PHM-NTS who received the bundled intervention of bedside delivery and parental education by the community pharmacist over the denominator of the number of infants discharged from the PHM-NTS that month. The outcome measure numerator was the number of newborns discharged from our PHM-NTS following up in our resident clinic for a 2-, 4-, or 6-month well child visit who were documented to be appropriately supplemented with vitamin D per parental report over the denominator of the number of newborns discharged from the PHM-NTS who were seen in our resident clinic. An infant may have been counted 1, 2, or 3 times depending on their enrollment date and attendance at 2-, 4-, and 6-month well child visits during our QI project period. Data were collected at monthly intervals to avoid double counting infants for each data point. Infants were tracked as part of this outcome measure at >1 visit if documented, allowing us to monitor ongoing adherence to this recommendation throughout the first 6 months of life and aid our outpatient colleagues in identifying patients struggling with obtaining refills or other potential barriers.

Some clinical staff voiced concern that bedside delivery of medication and teaching could delay discharge, thus the percentage of PHM-NTS discharges before 2 pm was tracked as our balancing measure and reviewed at monthly meetings.

Because many of our interventions were implemented concurrently, we were unable to study their effects individually. We used a custom-built EHR report to track the number of patients on the PHM-NTS with a vitamin D prescription at the time of discharge. These numbers were used in the calculation of our primary process measure. The community pharmacy tracked the total number of electronic prescriptions received for vitamin D from the nursery, as well as the number of patient encounters for bedside delivery and teaching by a pharmacist. Chart review was conducted for all patients who were discharged from the PHM-NTS and followed up for well visits in our pediatric resident clinic through the 6-month well child visit. This subset of patients was used in the calculation of our outcome measure. A previously created report of nursery discharges occurring by 2 pm was reviewed monthly as our balancing measure.

Although they overlap, our process measure baseline starts earlier than our outcome measure. This was related to difficulty with outpatient data collection, requiring outpatient EHR note template modification to make infant diet and medication data capturable elements.

Statistical process control p charts were used to display our process, outcome, and balancing measures. The control p charts were created by using Associates in Process Improvement rules for detecting special causes,¹⁹  a feature of QI Charts, a Shewhart control chart application for Microsoft Excel (PIP Products, Austin, Texas).²⁰ 

This project was submitted to the Maine Medical Center Institutional Review Board and determined to be QI and not research, and, as such, met the criteria for exemption from review.

Data from January 2016 to December 2016 served as preimplementation baseline for our process measure; during this period, 587 infants were discharged from the PHM-NTS. Our baseline data revealed a highly variable monthly mean of 7% to 33% with a cumulative mean of 24.8% for the year of newborns prescribed vitamin D and sent home with a bottle at discharge.

During the study period of January 2017 to December 2018, 1177 newborns were discharged from the PHM-NTS. Figure 3 displays improvement in our process measure over time with an interpretation of special cause points. Documentation of discharge from the PHM-NTS with the first bottle of vitamin D and teaching from a community pharmacist improved from 24.8% to 98%. Within 15 months, the process measure increased to and was sustained at a monthly mean of 98% for the subsequent 28 months. Points of special cause included (1) starting monthly interdisciplinary QI meetings and reviewing the importance of outpatient vitamin D supplementation with newborn staff (January 2017), (2) updating the written discharge instructions to include a section on the importance of vitamin D supplementation (May 2017), (3) creating a BPA (June 2017), and (4) converting the BPA to a standing vitamin D order in our newborn admission order set (December 2019).

During the baseline time period of September 2016 to December 2016, a mean of 49% of infants discharged from the PHM-NTS following up in our resident ambulatory clinic at their 2-, 4-, and 6-month well child visits, were documented as being appropriately supplemented with vitamin D per parental report.

Figure 4 displays improvement over time for our outcome measure with an interpretation of special cause points. Appropriate vitamin D supplementation as reported by parents at 2-, 4-, or 6-month well child visits increased from 49% to 89% for our pediatric resident clinic patients within 12 months of project initiation (1625 patient encounters). Points of special cause included (1) starting monthly interdisciplinary QI meetings and reviewing the importance of vitamin D supplementation with newborn staff and (2) increasing the rates of newborn discharge with the first bottle of vitamin D and teaching.

Figure 5 reveals the percentage of discharges from the PHM-NTS before 2 pm did not decrease during the study period. During the optimization and monitoring period, there was an increase in the percentage of infants discharged before 2 pm from 60% to 65% based on special cause variation. The increase was related to an increased need for postpartum beds that led to increased attentiveness to patient flow on the part of nursing and physician staff; however, it was not sustained and, therefore, no center line shift was made.

By discharging families with their first bottle of infant vitamin D oral solution along with increasing education for both families and providers, we were able to improve our infant supplementation rates over time as measured in the patient’s outpatient medical home. During the latter half of the baseline period, as we considered this QI project, we saw a Hawthorne effect. When we created an order set for vitamin D prescription and community pharmacy consultation for delivery and teaching during the development stage but had not yet formally introduced it, several providers in our division began to make use of the order set, although this was not consistent throughout the group. This drove home the importance of creating an interdisciplinary team and a key driver diagram to explore what components of the process needed to be addressed to ensure consistency. The interdisciplinary nature of our group was a particular strength of this project. By engaging lactation consultants, unit-based nursing educators, bedside nurses, and physicians we were able to sustain nursery-wide enthusiasm for this effort. Taking advantage of our onsite community pharmacy and the bedside delivery and teaching resource allowed us to use the newborn discharge phase of care to increase reported adherence to supplementation recommendations later in the first year of life. These results mirror our previous project on the general IPU. We hope this accomplishment will help our patients avoid short- and long-term sequelae of vitamin D deficiency.

As an intervention to promote nursery staff commitment and awareness, we encouraged all newborn providers, including family medicine and private pediatricians, to use the systems we developed and track the process measure for all newborns at our facility (Fig 3). We noticed some non-PHM providers in our nursery were not using the BPA or protocol order set. Through outreach and ongoing dialogue, including partnering with 1 private practice to develop maintenance of certification project, we achieved nursery-wide buy-in and participation in our intervention.

Implementing our project has been correlated with an improved rate of vitamin D supplementation for our PHM-NTS patients as measured via parental report in their outpatient medical home. Bedside delivery of medications, education, and teach-back have been shown to increase adherence to medication care plans,^{15–17,21–23}  which is consistent with our results. Our residency clinic population has a high proportion of patients who are insured by Medicaid and speak a language other than English primarily. We feel that, because our interventions were successful in this group, they could be replicated in other practices with a differently insured, primarily English-speaking population.

To our team’s knowledge, there are no specific publications that report improving vitamin D supplementation by creating an Inpatient Newborn Care/Community Pharmacy/Outpatient Primary Care interdisciplinary team. The vitamin D supplementation rate of 89% achieved and sustained in our pediatric resident clinic by the end of this project well exceed previously published national rates of 20% to 30% and our own baseline rate of 49%.^2–6 

Despite the success of our project, we had an admission for failure to thrive with a final diagnosis of vitamin D deficient rickets in a patient discharged during the study period. Despite our efforts at education and use of interpreters, the mother had a culturally based suspicion our strong messaging and efforts to provide a bottle of vitamin D for her infant were biased and potentially harmful. She did not administer the vitamin D as instructed for fear the bottle contained other components that would harm her infant. There are multiple approaches that could be taken to improve this situation or others that may arise, several of which have been undertaken at our institution. They include the creation of a Patient and Family Advisory Council (with attention to ethnic and cultural diversity within the council membership), ensuring the consistent use of interpreters, availability of educational materials translated into languages other than English, exploring the use of cultural brokers in situations in which there is disagreement between families and the medical team, trauma-informed care education for providers and staff, and other institutional diversity, equity and inclusion efforts.^24–28 

Limitations include the fact that we have different lengths of baseline data collection time frames for our process and outcome measures. Our project has limited generalizability due to the fact that our hospital has an on-site community pharmacy that was committed to this project and provided free vitamin D for families that did not have the ability to pay, a feature not available to all birthing hospitals, which may need to contract with a local pharmacy for medication delivery and find alternate sources of funding. In addition, we relied on parental reports of supplementation rather than bottle refill count or observed dose administration, which may have falsely elevated our outcome. We did not monitor for or address health disparities, and finally, our project was focused on a small portion of our nursery population and the results may not represent the entire population of our nursery.

Using the time of birth hospitalization to distribute a bottle of vitamin D and educate families regarding its administration and importance increased subsequent supplementation rates in the first 6 months of life. Educating the newborn care team about the importance of outpatient vitamin D supplementation and current AAP recommendations, updating our educational materials for families, and consistently dispensing a bottle of vitamin D, with teaching, before discharge from the nursery were all important steps in improving outpatient vitamin D supplementation for infants. These basic interventions could be replicated at other institutions.

Additional teaching for non-English speaking patients and attention to cultural attitudes and perspectives about vitamin D supplementation are future points of focus for our team.

We would like to thank Dr Leah Mallory for sharing lessons learned in care transitions, as well as for editorial assistance with this manuscript, our community pharmacists for providing the invaluable service of bedside medication delivery and teaching, and our nursery staff for providing education to families and essential feedback to the interdisciplinary QI group regarding process improvement.

AAP

American Academy of Pediatrics

BPA

best practice advisory

EHR

electronic health record

IPU

inpatient pediatric unit

PHM

Pediatric Hospital Medicine

PDSA

Plan-Do-Study-Act

PHM-NTS

Pediatric Hospital Medicine Newborn Teaching Service

QI

quality improvement