Pilot Study of Intrapartum Insulin Pump

Optimal intrapartum glucose management remains an unresolved challenge in the care of pregnant individuals with pre-gestational diabetes, despite its critical importance to maternal and neonatal outcomes.1 Approximately 18 million women in the United States have diabetes mellitus2, and pregestational diabetes complicates 1-2% of all pregnancies3. Poor glycemic control during labor is associated with an increased risk of neonatal hypoglycemia, respiratory distress, hyperbilirubinemia, and dangerous maternal glycemic variability.1 However, best practices for intrapartum glucose management remain inadequately defined and warrant further investigation.4,5,6 Automated insulin delivery (AID) systems and modern continuous subcutaneous insulin infusion (CSII) technology now allow for precise glucose management and have transformed diabetes care, emerging as a new standard for glycemic management during pregnancy. Despite routine use in the antenatal period, these pumps are usually discontinued upon hospital admission for labor with transition to more traditional intravenous (IV) insulin infusions.7 This may cause glycemic instability, is labor-intensive, burdens both patients and clinical teams,4 and diminishes patient autonomy.8,9 Urgently addressing the gap in clinical evidence to guide CSII/AID use during labor could change clinical practice, optimize glucose management protocols during labor, and ultimately reduce preventable complications for both mother and baby.

Washington, DC provides an ideal environment to study innovations in intrapartum diabetes care. Diabetes prevalence reaches 19.7% in underserved wards10,11, much higher than the national rate of 11.6%. MedStar Health is a major regional network serving a broad and diverse patient population. Within this system, the MedStar Diabetes in Pregnancy Program delivers multidisciplinary care to hundreds of pregnant individuals with diabetes each year. With robust clinical infrastructure, strong community partnerships, and a focus on reducing barriers to care, our team is ideally positioned to evaluate CSII/AID therapy during labor in a real-world, high-risk population.

The overall objective of this pilot study is to evaluate the feasibility, safety, and acceptability of CSII during labor. This will provide foundational data to inform the design of a potentially practice and guidelines-changing multicenter randomized controlled trial (RCT). Our long-term goal is to determine whether CSII is a safe and effective alternative to IV insulin, the current standard of care, for intrapartum diabetes management. We hypothesize that use of insulin pumps in labor is safe, maintains glycemic control equivalent to or better than IV insulin, reduces neonatal hypoglycemia, and improves patient birth experience while reducing healthcare burden. To achieve this, we propose the following specific aims:

Specific Aim 1 Feasibility:[MM1.1] To assess the feasibility of intrapartum insulin pump management We will conduct a quasi-experimental, non-randomized design consisting of two arms: a prospectively enrolled cohort managed with CSII/AID during labor and a matched historical cohort managed with IV insulin. We will assess key feasibility metrics including insulin pump protocol creation and adherence, integration into clinical workflows, and recruitment and retention rates. We will evaluate preliminary effectiveness and safety by comparing maternal glycemic and neonatal outcomes between cohorts. These data will inform the design, protocol refinement, implementation strategies, and power calculations for a future multicenter trial.

Specific Aim 2 Stakeholder experiences: To characterize patient, provider, and nursing perspectives with intrapartum CSII/AID use.

Using mixed methods, we will assess patient, provider, and nurse-reported outcomes. Patient metrics will include satisfaction, perceived autonomy, and treatment burden. Providers and nursing staff will be surveyed and interviewed to evaluate acceptability, perceived barriers, and facilitators to implementation. Healthcare provider input will be utilized to create inpatient pump dashboards with Information Technology. Exploratory analyses will assess healthcare utilization and cost implications.

Together, these aims will generate essential preliminary data to inform the design of a future multicenter RCT evaluating CSII/AID therapy as an innovative approach to intrapartum glucose management. This work directly supports the mission of the Latham Fund by addressing a critical clinical gap in a high-risk, medically underserved population with a high prevalence of diabetes. With its potential to reduce preventable maternal and neonatal complications, enhance the birth experience for individuals with chronic disease, and improve access to patient-centered care, this study would be the key first step to advancing evidence-based practice and transforming intrapartum diabetes care for a vulnerable population.

Specific Aim 1 (Feasibility): To evaluate the feasibility of insulin pump management during labor and delivery. We will conduct a quasi-experimental, non-randomized design consisting of two arms: a prospectively enrolled cohort managed with CSII/AID during labor and a matched historical cohort managed with IV insulin. We will assess key feasibility metrics including insulin pump protocol creation and adherence, integration into workflows, and recruitment and retention rates. We will evaluate preliminary effectiveness and safety by comparing maternal glycemic and neonatal outcomes between cohorts. Careful attention will be paid to initial trends in frequency of adverse safety events: severe maternal hyper- and hypoglycemia, incidence of diabetic ketoacidosis, and syncopal events. These data will inform the design, protocol refinement, implementation strategies, and power calculations for a future multicenter trial.

Specific Aim 2 (Stakeholder experiences): To assess patient, provider, and nursing perspectives on insulin pump therapy versus IV insulin during labor.

We will collect qualitative and quantitative data on patient-reported outcomes such as satisfaction, sense of autonomy, and perceived treatment burden. Patient metrics will include satisfaction, perceived autonomy, and treatment burden. Providers and nursing staff will be surveyed and interviewed to evaluate acceptability, perceived barriers, and facilitators to implementation. Healthcare provider input will be utilized to create inpatient pump dashboards with Information Technology (IT). Exploratory analyses will assess healthcare utilization and cost implications. Additionally, we will evaluate provider and nursing satisfaction and comfort with each insulin delivery approach to understand facilitators and barriers to implementation.

OVERALL STUDY DESIGN We propose a quasi-experimental non-randomized pilot study to evaluate the feasibility, safety, and acceptability of continuing continuous subcutaneous insulin infusion (CSII) during labor in pregnant individuals with Types 1 and 2 diabetes mellitus. Participants will be prospectively enrolled and continued on personal insulin pump therapy (CSII with or without AID) throughout labor. We will compare to a historic cohort of antenatal CSII use transitioned to IV insulin during labor. Outcomes will focus primarily on feasibility metrics of CSII use intrapartum to allow refinement of protocols and implementation strategies for a future RCT. We will also evaluate initial trends in clinical safety outcomes (maternal and neonatal), and patient/provider-reported measures.

PARTICIPANTS AND SETTING The study will enroll individuals with a pre-pregnancy diagnosis of T1DM or T2DM presenting for delivery at MedStar Health, with antenatal care through our MedStar Diabetes in Pregnancy Program. Participants will have been on CSII (with or without AID) for at least 2 months prior to delivery admission.

The MedStar Diabetes in Pregnancy program is centered at Washington Hospital Center, a major regional hospital, serving as both a referral center for patients with complex pregnancies and a safety-net provider for historically underserved residents. Within this setting, the MedStar Diabetes in Pregnancy Program provides multidisciplinary care to hundreds of pregnant individuals with type 1 and type 2 diabetes annually. Our program serves patients primarily in Washington, DC with embedded outreach and social services through the Safe Babies, Safe Moms program. Because of the clinical success of our program, we have extended services to the MedStar South region, providing pregnancy diabetes care to patients delivering at our urban, academic hospitals (Washington Hospital Center, Georgetown) as well as to underserved urban and rural hospitals in the greater DC region (Southern Maryland Hospital Center, St. Mary's Hospital).

This pilot study is primarily designed to assess feasibility outcomes, including recruitment rate, protocol adherence (i.e., percentage of participants who successfully remain on insulin pump therapy throughout labor), and integration into clinical workflows. Based on institutional delivery volumes and diabetes prevalence, we anticipate enrolling 30 participants into the prospective intervention arm.

Inclusion criteria will include: singleton gestation, established T1DM or T2DM with use of CSII for ≥2 months prior to delivery, demonstrated cognitive ability to self-manage insulin, familiarity with pump technology, availability of necessary supplies, stable clinical status, and ability to provide informed consent.

Exclusion criteria will include planned cesarean delivery without labor, significant comorbidities affecting insulin requirements (e.g., steroid therapy), stillbirth prior to presenting for delivery, and/or known severe fetal anomalies.

INTERVENTIONS Participants in the intervention group will continue using their personal insulin pumps throughout the intrapartum period, under a standardized inpatient protocol developed specifically for this study. This protocol will be created in collaboration with endocrinology, Maternal Fetal Medicine, nursing leadership, and clinical informatics, with the goal of safely integrating outpatient diabetes technology into labor and delivery care.

As part of this effort, we will develop a structured insulin pump titration protocol, including:

• Glycemic targets for labor
• Criteria for dose adjustments or override
• Clear transition points for conversion to IV insulin if needed
• Nursing documentation procedures and safety checklists

In parallel, we will begin preparations to build a custom inpatient dashboard that integrates real-time glucose data and pump use status into the electronic health record (EHR), utilizing input from our healthcare stakeholders, including physicians and nurses. This dashboard will ultimately serve as a centralized tool for nursing, obstetric, and endocrinology teams to track glucose trends, ensure protocol adherence, and flag safety concerns in real time. It will also support clinical decision-making and facilitate oversight in the Labor and Delivery unit.

Participants will be recruited and screened for enrollment antenatally by study personnel. Upon hospital admission for labor induction or in early spontaneous labor, participants will be re-screened for eligibility to continue pump use. Inclusion will require demonstrated cognitive ability to self-manage insulin, familiarity with pump technology, availability of necessary supplies, and stable clinical status. Participants will self-administer insulin with oversight from the inpatient care team. Glycemic assessment will be via continuous glucose monitoring (CGM) with periodic capillary blood glucose level assessments for concordance. Nursing staff will be trained on the insulin pump protocol and safety criteria. The clinical and research care team will oversee the care of these participants.

Participants who become clinically unstable (altered mental status, persistent hyper/hypoglycemia, or unanticipated cesarean delivery under general anesthesia) will be transitioned to standard intravenous (IV) insulin therapy. All such transitions will be documented as part of adherence and safety tracking. This data will inform whether specific clinical factors may identify individuals at higher risk of adverse events, thereby guiding future eligibility criteria for an RCT.

The control group will consist of historical controls who were managed under prior standard care, which involved discontinuation of insulin pumps and initiation of IV insulin therapy during labor.

This pragmatic intervention is designed to test the real-world feasibility, safety, and acceptability of continuing insulin pump therapy during labor. By embedding this protocol and dashboard within existing inpatient systems, we aim to facilitate future scale-up and implementation across other hospital settings.

DATA COLLECTION AND MEASURES

• Aim 1 - Feasibility Metrics:
• Recruitment rate, consent rate, participant demographics, protocol adherence (percent of time on assigned insulin regimen), proportion that require transition to IV insulin, and data completeness.
• Maternal and Neonatal Safety Outcomes: [MM2.1]
• In addition to feasibility outcomes, we will collect secondary clinical outcomes. These outcomes will not be used for hypothesis testing but will provide preliminary estimates of effect sizes and variability to inform sample size calculations and outcome selection for a future multicenter RCT.
• Maternal: Time in target glucose range during labor (70-140 mg/dL), number of hypoglycemic (<70 mg/dL) and hyperglycemic (>200 mg/dL) events, incidence of diabetic ketoacidosis, need for intervention to correct abnormal blood glucose level
• Neonatal: Incidence of neonatal hypoglycemia (<45 mg/dL), respiratory distress, hyperbilirubinemia requiring treatment, neonatal intensive care unit admission, and birthweight outcomes.
• Aim 2- Stakeholder experience:
• Patient: Postpartum survey assessing satisfaction with intrapartum diabetes management, sense of autonomy, perceived burden, and willingness to use the assigned method in future deliveries.
• Provider/Nursing: Structured surveys on comfort, workload, and perceived safety of insulin management protocol.
• Healthcare system: Cost-effectiveness analysis to evaluate potential cost savings associated with insulin pump therapy compared to intravenous insulin. This will include analysis of direct costs such as insulin delivery (pump vs. IV drip), pharmacy resource utilization, and nursing time or support required for each approach.

DATA ANALYSIS PLAN As a pilot study, this trial is not powered for hypothesis testing. Descriptive statistics will summarize feasibility outcomes (Aim 1) with point estimates (means, standard deviations, quantiles for continuous variables and frequencies and percentage proportions for categorical variables) compared against predetermined feasibility criteria. Assessment of recruitment, consent, attrition rates, and preliminary estimates of sample size and uncertainty will be utilized to calculate sample size for future trial.

To assess recruitment feasibility, we will estimate the proportion of eligible participants who enroll in the study. Assuming a recruitment rate of 75%, a sample size of 30 participants will allow us to estimate this rate with a 95% confidence interval (CI) of approximately ±15% (CI range: ~60%-90%).

We will also evaluate protocol adherence, defined as the proportion of participants who successfully remain on insulin pump therapy throughout the intrapartum period per protocol. If we assume an adherence rate of 85%, our sample size will yield a 95% CI of approximately ±13% (CI range: ~72%-98%). These precision estimates are consistent with recommended benchmarks for pilot studies and will guide refinement of intervention protocols, implementation strategies, and sample size calculations for a future multicenter randomized controlled trial.

Intervention participants (n=30) will be compared in a 2:1 ratio to historical controls (n=60; total n=90) who received intravenous insulin during labor. These exploratory comparisons will examine maternal glycemic control, neonatal outcomes, and health system impacts.

For clinical safety outcomes, we will categorize and report all adverse events, serious adverse events, and any unexpected events. We will assess time-in-range metrics, estimate rates and variability of maternal and neonatal outcomes through point estimates with standard deviations, and preliminary hypothesis testing results through analysis of outcome variability.

For Aim 2, patient and provider-reported outcomes will be summarized using descriptive statistics. Qualitative data will be analyzed thematically using standard coding procedures. Feedback will be utilized to inform procedures in the larger future trial.

This project represents a critical step in establishing my research niche at the intersection of diabetes and cardiovascular disease management during pregnancy. As an early-career investigator, my goal is to develop a sustained program of clinical research that improves care paradigms for individuals with pregestational and gestational diabetes, while expanding access to high-quality, multidisciplinary care through MedStar's Diabetes in Pregnancy program.

The findings from this pilot study will directly inform and serve as preliminary data for a fully powered, multisite randomized controlled trial aimed at evaluating the clinical effectiveness, safety, and acceptability of continuing CSII during labor. Importantly, this future study will be designed to assess maternal and neonatal outcomes with sufficient power to influence clinical practice guidelines.

Pending successful completion of the pilot, we plan to pursue funding through the NICHD, NIDDK, or targeted foundation grants or partnerships with diabetes technology companies to support a multicenter trial. This subsequent study will involve collaboration with additional academic medical centers to ensure geographic and demographic diversity, improve generalizability and sample size, and engage broader provider perspectives. We will also consider partnering with professional organizations such as The American College of Obstetrics and Gynecology (ACOG), Society for Maternal-Fetal Medicine (SMFM), and the American Diabetes Association (ADA) to support dissemination and guideline development.

Long-term, this line of research has the potential to:

• Establish evidence-based recommendations for intrapartum glucose management using CSII.
• Promote patient-centered models of diabetes care during labor.
• Reduce maternal and neonatal complications associated with suboptimal glycemic control.
• Decrease healthcare system burden through streamlined insulin management protocols.