Between-visit care represents a substantial and underrecognized component of health care utilization in pediatric inflammatory bowel disease

Abstract

Background

Children with inflammatory bowel disease (IBD) require substantial health care engagement, yet traditional utilization metrics underestimate the scope of care by omitting between-visit clinical activity such as electronic messaging, telephone encounters, and results review. Pediatric-specific data describing this hidden workload remain limited.

Methods

We conducted a retrospective longitudinal cohort study of youth aged 0 to 22 years diagnosed with IBD between 2015 and 2024 at a tertiary care center. Eligible patients had ≥ 6 months of primary care prior to diagnosis and ≥ 6 months of gastroenterological follow-up postdiagnosis. From the shared electronic health record, we quantified outpatient visits, emergency care, and procedures, as well as between-visit care (portal messages from patients, telephone encounters, and clinician-patient review of results) from prediagnosis through long-term follow-up. Events were aggregated into 6-month epochs relative to diagnosis and stratified by disease severity using validated symptom indices. Predictors of between-visit care were evaluated using generalized linear mixed-effects modeling.

Results

Among 226 patients (median age 12 years; 57% male), health care utilization increased markedly during peri-diagnostic epochs. Between-visit care rose from a median of 1 event per 6 months > 1 year prediagnosis to 64 events in the year following diagnosis, and the number of these events remained persistently elevated thereafter. Higher disease severity was strongly associated with increased between-visit care. In multivariable analysis, postdiagnosis epochs, moderate-severe disease activity, and later diagnostic year independently predicted increased between-visit care (all P < .0001).

Conclusions

Between-visit care represents a substantial and clinically meaningful component of pediatric IBD management, underscoring the need for workforce planning, reimbursement models, and digital innovations to address this largely invisible clinical workload.

Brief Summary

Between-visit care is a substantial and persistent component of pediatric inflammatory bowel disease management, with the frequency increasing sharply at diagnosis and remaining elevated over time. Quantifying this hidden workload is essential for workforce planning, reimbursement reform, and sustainable and quality care delivery models.

Key Messages.

What is already known? 

Between-visit communication contributes significantly to workload in adult inflammatory bowel disease (IBD) care, but pediatric-specific data are limited.

What is new here? 

This longitudinal pediatric study comprehensively quantifies between-visit care from prediagnosis through long-term follow-up, demonstrating that the between-visit care burden rises sharply at diagnosis and remains persistently elevated. Higher between-visit care volume is also associated with greater disease severity.

How can this study help patient care? 

By making invisible clinical work measurable, these findings support accurate workforce planning, inform reimbursement reform, and guide development of digital and self-management tools to sustain high-quality pediatric IBD care.

Introduction

Patients with inflammatory bowel disease (IBD), both adults and children, experience substantially higher health care utilization than non-IBD counterparts.¹ Prior evaluations have largely focused on discrete medical encounters, including outpatient clinic visits, endoscopic and surgical procedures, emergency department (ED) visits, and hospitalizations.^1–3 While these metrics describe important facets of care, they do not fully capture the breadth of clinical activity required to manage this chronic condition.

A significant portion of IBD care occurs between in-person visits. These between-visit interactions include telephone encounters, electronic messages, clinician review of laboratory, imaging, and pathology results. Such work is clinically necessary yet often underrecognized and unreimbursed. For example, 1 adult gastroenterology practice reported nearly $170 000 annually in costs associated with medication-related phone calls alone.⁴ Additionally, increased between-visit communication has been shown to predict future health care utilization,⁵ suggesting that these interactions reflect genuine clinical need rather than discretionary contact. In pediatric subspecialty care, for which team-based communication, caregiver involvement, and developmental considerations are integral,⁶ the true magnitude of this between-visit workload may be even greater.

Historically, telephone calls were the primary means of communication between families and clinicians. With the widespread adoption of electronic health records (EHRs) and patient portals, asynchronous digital messaging has become increasingly common. Although these platforms may enhance access and patient engagement, evidence suggests they do not simply replace telephone encounters; instead, they may increase overall communication volume.⁷ Lower thresholds for initiating messages, the need for iterative clarification, and miscommunication may all contribute to this rise, raising concerns about added workload and potential clinician burnout. Yet, few studies have examined how these communication patterns evolve across the diagnostic trajectory or how they relate to disease activity over time.

Metadata from the EHR portal provide a scalable opportunity to quantify the scope of between-visit care. However, most existing data examining portal use and between-visit workload have focused on adults with IBD. Adult studies have revealed substantial variability in portal messaging volume,⁸ yet comparable data in pediatric gastroenterology remain scarce. Pediatric IBD care introduces unique dynamics⁶—including the involvement of caregivers, evolving patient autonomy, and heightened expectations regarding access to the clinical team—suggesting that communication patterns may differ meaningfully from those observed in adults.

Understanding the timing, magnitude, and predictors of between-visit communication is essential to improving care delivery and accurately estimating workforce needs to effectively manage pediatric IBD patients. To address these knowledge gaps, we sought to comprehensively quantify total health care utilization—including between-visit care—from the prediagnosis period through long-term follow-up among pediatric patients with IBD. To address these gaps, we sought to longitudinally quantify total health care utilization—including traditionally invisible between-visit care—from the prediagnosis period through extended follow-up, and to identify clinical and temporal predictors of communication burden in pediatric IBD.

Materials and methods

The primary objective of this study was to comprehensively quantify health care utilization, including between-visit care, using EHR data from the prediagnosis period through long-term follow-up among pediatric patients with inflammatory bowel disease. The secondary objective was to evaluate whether health care utilization patterns differed by disease severity, as measured by validated symptom scores.

We identified pediatric patients diagnosed with IBD at a tertiary care center during 2015-2024. Eligible patients had an activated EHR portal account and received primary care within an affiliated network for ≥6 months prior to diagnosis as well as subspecialty care within the gastroenterology subspecialty service for ≥6 months after diagnosis, ensuring consistent capture of all encounters. Using the shared EHR, we extracted all health care events, including outpatient visits, ED encounters, hospitalizations, procedures, telephone encounters, portal messages from patients, and clinician-patient review of laboratory and diagnostic results.

Demographic data extracted included age, sex, race, ethnicity, and IBD subtype. Disease activity was assessed using validated indices: the Pediatric Ulcerative Colitis Activity Index⁹ (PUCAI) for ulcerative colitis and the weighted Pediatric Crohn’s Disease Activity Index¹⁰ (wPCDAI) for Crohn disease. Clinical symptom data were captured at clinical encounters (both via telemedicine and in person), which occured every 3 to 6 months.

All health care events (including outpatient visits, ED encounters, hospitalizations, procedures, telephone encounters, portal messages from patients, and clinician-patient review of laboratory and diagnostic results) were counted individually. If an outpatient encounter led to an ED encounter which resulted in a hospitalization and/or procedure, all 4 encounters were tallied; similarly, if a portal message led to additional messaging, each message was counted independently. Events were aggregated into 6-month epochs and classified relative to diagnosis (>1 year prior, 1 year prior, 1 year after, and >1year after diagnosis). Event rates were also evaluated according to disease severity (asymptomatic, mild, or moderate-severe) based on clinical symptom scores. For utilization outcomes, absence of events within an eligible follow-up period was recorded as zero. Missing clinical variables (eg, symptom scores) were not imputed, and analyses were restricted to observed values.

Median and IQR values were reported for continuous variables. Group differences were assessed with Kruskal-Wallis testing given non-normal data distributions. A generalized linear mixed model (Table 4) was used to identify independent predictors of between-visit communication volume (which was transformed to a normal scores distribution through ranking), with the patient identifier included as a random effect and all other covariates specified as fixed effects. Specific fixed-effect variables evaluated included age at time of event, year of diagnosis, patient sex, patient race (White vs non-White), patient ethnicity (Hispanic vs non-Hispanic), preferred language (English vs non-English), disease subtype (Crohn vs ulcerative colitis/indeterminate colitis), event time–based grouping, and disease severity (moderate-severe vs other).

Table 4.

Generalized linear mixed-effects model predicting between-visit care volume.^a

Variable	Estimate	IQR	P value
Ethnicity (Hispanic vs non-Hispanic)	−0.10	(−0.24 to 0.05)	.18
Sex (female vs male)	−0.10	(−0.22 to 0.03)	.12
Race (non-White vs White)	0.002	(−0.14 to 0.14)	.98
Age at event, y	0.01	(−0.01 to 0.02)	.49
Disease subtype (ulcerative colitis/indeterminate colitis vs Crohn)	0.03	(−0.10 to 0.17)	.63
Preferred language (English vs non-English)	0.16	(−0.13 to 0.44)	.28
Disease severity (moderate-severe vs asymptomatic/mild)	0.15 b	(0.12-0.18)	<.0001
Year of diagnosis	0.35	(0.27-0.42)	<.0001
Time epoch (after vs before IBD diagnosis)	0.78	(0.67-0.89)	<.0001

Abbreviation: IBD, inflammatory bowel disease.

^a $\chi$ ² = 170.1. Whole model < 0.0001.

^bBoldface type indicates statistical significance.

^cPatient identifier entered to control for random effects.

Results

In total, 226 youth aged 0 to 22 years were included. Demographic data are presented in Table 1. Overall, the cohort of included patients was primarily male (57.0%), White (66.4%), and non-Hispanic (71.2%), with a median age at diagnosis of 12 years. A majority (59.7%) of patients had Crohn disease with a median follow-up period of 3.0 years.

Table 1.

Demographic and clinical data of the study cohort (n = 226).

Variables	Values
Age at diagnosis, y, median (IQR)	12 (9-4)
Sex, male: female, n	129: 97
Race, n	White, 150 Black, 13 Asian, 24 Hispanic, 24 Middle Eastern/North African, 8 Decline to answer, 6 Other, 1
Ethnicity, Hispanic: non-Hispanic: decline, n	62: 161: 3
Preferred language, English: other, n	215: 11
Disease subtype, Crohn: indeterminate colitis: ulcerative colitis, n	135: 32: 59
Paris Classification	Crohn (n = 135)
	Age: A1a-43; A1b-86; A2-6  Location: L0-4; L1-18; L2-43; L3-70; L4a-51; L4ab-5; L4b-10  Behavior: B1-124; B2-8; B3-3  Perianal: with-14;without-121  Growth delay: with-39; without-96
	Colitis (n = 91)
	Age: A1a-14; A1b-77  Extent: E1-4; E2-13; E3-5; E4-69  Severity: S0-42; S1-49
Year of diagnosis, median (IQR)	2021 (2019-2023)
Duration of follow-up following diagnosis, y, median (IQR)	3.0 (1.6-5.1)

Health care utilization increased substantially over the time epochs analyzed among the pediatric patients with IBD (Table 2 and Figure 1). Emergency health care utilization—particularly ED encounters and hospitalizations—increased significantly in the peri-diagnosis epochs (within 1 year before and after diagnosis) and subsequently declined. Regarding ambulatory care, visit volume shifted from primary care to the subspecialty service, with gastroenterology visits increasing sharply during the first year after diagnosis and then stabilizing although the visit volume remained higher than that prior to diagnosis. Procedural volume also increased during the peri-diagnosis epochs in particular and then subsided after the first year after diagnosis but remained higher than the prediagnosis volume. Between-visit care events (patient-initiated portal messages from patients, results reviewed, and telephone encounters) similarly increased during the first year following diagnosis, reaching a median of 64 (IQR, 29-123) events per 6 months, with persistence at lower levels thereafter (median 41, IQR 17-71 events per 6 months), although event frequencies were notably higher than prior to diagnosis (1 [0-6] and 8 [1-37] events per 6 months, > 1 year prior to diagnosis and in the year prior to diagnosis respectively). Collectively, between-visit care accounted for a substantial proportion of total health care events exceeding the volume of scheduled outpatient visits in postdiagnostic epochs, as shown in Figure 1. Unlike emergency and procedural utilization, which declined after the first postdiagnostic year, between-visit care demonstrated sustained elevation across long-term follow-up.

Table 2.

Health care utilization by time epochs relative to IBD diagnosis.

Health care event	Time epoch				P valuea
	>1 y Prior to IBD diagnosis (n = 180)	<1 y Prior to IBD diagnosis (n = 226)	<1 y After IBD diagnosis (n = 226)	>1 y After IBD diagnosis (n = 174)
Ambulatory care, median (IQR)
Primary care visits	1 (0-3)	1 (0-4)	0 (0-2)	0 (0-2)	<.0001
Subspecialty care vsits	0 (0-0)	0 (0-1)	2 (1-3)	1 (1-2)	<.0001
Procedures,, median (range)
Endoscopy procedures	0 (0-2)	0 (0-3)	0 (0-3)	0 (0-5)	<.0001
Emergency care, median (range)
ED visits	0 (0-4)	0 (0-4)	0 (0-5)	0 (0-3)	<.0001
Hospital admissions	0 (0-5)	0 (0-3)	0 (0-3)	0 (0-4)	<.0001
Between-visit care, median (IQR)
Portal messages from patients	0 (0-3)	3 (0-12)	30 (11-56)	22 (7-42)	<.0001
Telephone encounters	0 (0-0)	0 (0-17)	25 (6-58)	11 (2-26)	<.0001
Clinician-patient results review	0 (0-1)	2 (0-5)	6 (3-11)	4 (2-7)	<.0001

Abbreviation: ED, emergency department; IBD, inflammatory bowel disease.

^a P value refers to nonparametric (Kruskal-Wallis) testing results.

Figure 1.

Health care utilization events per 6-month period relative to inflammatory bowel disease (IBD) diagnosis among 226 pediatric patients. Individual points represent observed event counts, with overlying boxplots indicating median and IQR. Events include ambulatory care, emergency care, procedures, and between-visit care. Patients could contribute observations to multiple categories across time periods. See Table 2 for exact values and additional details. ED,  emergency department; GI, gastroenterology.

In general, health care utilization increased in parallel with worsening disease severity (Table 3 and Figure 2). In particular, subspecialty care visits and emergency care visits increased with worsening disease severity. Procedural events did increase in patients with mild disease in particular compared with patients with moderate-severe symptoms and more so than those without symptoms. Between-visit care volume also increased significantly with worsening disease activity, reaching a median of 84 (IQR, 40-138) events per 6 months among patients with moderate-severe symptoms compared to asymptomatic patients (48 [IQR, 28-72] events per 6 months) and those with mild symptoms (56 [IQR, 30-96] events per 6 months). In contrast, primary care visits did not change across disease severity levels. Notably, while emergency and procedural utilization increased with worsening symptoms, primary care visit frequency did not vary by disease severity, suggesting that escalating clinical need was largely managed within subspecialty and between-visit care pathways.

Table 3.

Health care utilization by disease severity relative to IBD diagnosis.

Health care event	Disease severity			P value
	Asymptomatic (n = 144)	Mild (n = 148)	Moderate-severe (n = 134)
Ambulatory care, median (IQR).
Primary care visits	0 (0-2)	0 (0-2)	0 (0-2)	.25
Subspecialty care visits	1 (1-2)	2 (1-2)	2 (1-3)	<.0001
Procedures, median (range)
Endoscopy Procedures	0 (0-2)	0 (0-5)	0 (0-2)	.0002
Emergency care, median (range)
ED visits	0 (0-5)	0 (0-3)	0 (0-4)	<.0001
Hospital admissions	0 (0-2)	0 (0-2)	0 (0-4)	<.0001
Between-visit care, median (IQR)
Portal messages from patients	27 (14-43)	30 (15-49)	39 (14-67)	.0002
Telephone encounters	3 (2-6)	5 (3-8)	8 (4-12)	<.0001
Clinician-patient results review	14 (5-26)	19 (7-39)	30 (11-69)	<.0001

Abbreviation: ED, emergency department.

^a P values refer to nonparametric (Kruskal-Wallis) testing results.

Figure 2.

Health care utilization events per 6-month period by inflammatory bowel disease (IBD) disease activity state based on validated clinical symptom scores among 219 pediatric patients with available disease activity data. Individual points represent observed event counts, with overlying boxplots indicating median and IQR. Patients could contribute observations to multiple categories across disease activity states. See Table 3 for exact values and additional details. ED, emergency department; GI, gastroenterology; Mod-Severe, moderate to severe.

In multivariable analysis (Table 4), time epochs after diagnosis, moderate-severe symptoms, and year of diagnosis were strong predictors of increased between-visit care (all P < .0001). Demographic factors were not associated with between-visit care volume. The absence of associations with demographic characteristics suggests that between-visit communication burden is driven primarily by clinical and temporal factors rather than patient sociodemographic differences.

Discussion

In this longitudinal study of pediatric patients with IBD, we comprehensively quantified health care utilization across the disease course, with particular emphasis on between-visit care. Our findings demonstrate that between-visit care constitutes a substantial and sustained component of pediatric IBD management that increases sharply around diagnosis and remains persistently elevated during long-term follow-up. Importantly, this communication burden parallels disease activity and is independently associated with postdiagnosis time, greater symptom severity, and more recent diagnostic years highlighting its clinical relevance and evolving nature.

A central finding of this study is the magnitude and durability of between-visit care. While increases in outpatient visits, emergency encounters, and procedures were most pronounced during peri-diagnostic epochs with subsequent decline, between-visit care remained persistently elevated well beyond the first year following diagnosis. Even during periods of relative disease quiescence, patients generated substantial volumes of portal messages, telephone encounters, and result review requirements. These findings underscore that the longitudinal management of pediatric IBD extends far beyond scheduled clinic visits and episodic care. This persistent care burden likely reflects the complexity of pediatric IBD management including medication titration, symptom monitoring, laboratory surveillance, comorbidity screening, insurance navigation, school-related concerns, anticipatory guidance, administrative needs, transitional readiness planning, and caregiver and patient education.¹¹ Unlike acute care utilization, which may diminish as disease control improves, the need for ongoing communication and care appears intrinsic to chronic disease management in the pediatric IBD population.

We observed a strong, graded relationship between disease severity and health care utilization, with the greatest increases seen in subspecialty visits, emergency care, and between-visit care among patients with moderate to severe symptoms. This association between disease severity and between-visit care aligns with prior adult studies demonstrating that higher messaging volume predicts future flares and health care utilization, functioning as an early indicator of disease exacerbation or unmet clinical needs.⁵ Rather than discretionary contact, these interactions appear to represent timely requests for care during vulnerable periods. Future work may explore whether between-visit care patterns can be leveraged for earlier intervention or targeted support.

The year of diagnosis has emerged as an independent predictor of increased between-visit care, with later diagnostic years associated with higher communication volume. This finding likely reflects broader trends in digital health adoption,¹² including increased reliance on patient portals as a primary communication modality. Importantly, our data, correlating with findings of others,⁷ suggest that portal messages from patients did not replace telephone encounters; rather, both modalities increased following diagnosis, contributing cumulatively to clinician workload.

The expansion of asynchronous communication may lower barriers to access and enhance engagement for families, particularly during vulnerable periods. However, this increase also introduces new workload demands often unaccounted for in traditional staffing models¹³ and reimbursement frameworks. These findings highlight a growing mismatch between care delivery expectations and the resources allocated to support them.

We observed modest but notable emergency health care utilization in the years preceding IBD diagnosis, suggesting that some children experience escalating symptoms that prompt acute care encounters prior to definitive diagnosis. This finding aligns with a meta-analysis of 24 studies demonstrating 2-24 month delays in the diagnosis of IBD in youth.¹⁴ These prediagnostic encounters represent potential missed opportunities for earlier recognition and referral, with implications for symptom burden, health care costs, and long-term outcomes.

Despite its clinical importance, between-visit care remains largely invisible in traditional utilization metrics and is frequently unreimbursed. Our findings indicate that this hidden workload is substantial, persistent, and closely tied to disease complexity rather than patient demographic factors. As health care systems increasingly emphasize digital access and patient engagement, failure to recognize and support this work risks exacerbating clinician burnout and undermining care quality. Previous work suggests that unreimbursed communication cost a gastroenterology practice over $19 000 per month,⁴ not including indirect impacts such as workflow disruption and clinician fatigue. This burden exists alongside already existing substantial EHR documentation time within visits.¹⁵

Recent reimbursement pathways for asynchronous care represent an important step forward, but their applicability and sustainability in pediatric subspecialty care remain uncertain. Quantifying between-visit care, as we have done here, provides a necessary foundation for more equitable workforce planning and payment models that reflect the realities of chronic disease management.

Emerging digital health tools, including artificial intelligence (AI)–assisted messaging and automated triage systems, offer promise in mitigating between-visit workload. Early studies suggest that AI-generated draft responses to patient messages may reduce perceived cognitive burden and exhaustion among clinicians.¹⁶ However, current evidence does not yet demonstrate objective time savings, and questions remain regarding message clarity, appropriateness, and equity, especially for families with lower health literacy.¹⁷ While AI has great promise to collate information and potentially interpret information patterns that may help with clinical decision making, trust¹⁸ remains a major barrier and issue with its adoption given concerns with medical liability¹⁸ as well as demonstrated issues with data input misinterpretation and hallucinations.¹⁹ Careful validation and thoughtful implementation will be critical before these tools can be relied upon to meaningfully reduce workload.

Promoting disease self-management is a core goal of pediatric IBD care, particularly as youth transition to adult care. Interestingly, our experience suggests that efforts to enhance autonomy—such as transition readiness education—may initially increase portal engagement (Gilligan et al. unpublished data) as adolescents assume greater responsibility for their health.²⁰ This transient rise in communication should be viewed as a developmental investment rather than a failure of efficiency, reinforcing the need for balanced expectations when evaluating patient-health care team communication volume.

This study has limitations. As a single-institution analysis, findings may reflect local practice patterns, although the integrated health care system served by a single EHR enabled comprehensive capture of utilization across care settings. Also, inclusion of both phone and electronic modalities strengthens the comprehensiveness of our assessment. We did not differentiate communication by provider role or message complexity, which may further refine understanding of workload. Additionally, while disease activity was assessed using validated indices, symptom data were only available reliably at observed encounters and not imputed between visits. Nevertheless, findings are concordant with reports from adult gastroenterology practice,^5,8 supporting generalizability.

Future studies should evaluate the clinical content, urgency, and downstream outcomes associated with between-visit communications to distinguish routine management from early indicators of disease exacerbation. Multicenter studies are needed to assess generalizability across practice settings and to examine provider-level and health system contributors to communication burden. Prospective evaluation of digital triage tools, AI-assisted message support, and structured self-management interventions will be critical to determine whether these approaches can sustainably reduce clinician workload without compromising access, equity, or quality of care. Finally, integration of between-visit care metrics into value-based reimbursement models represents an important area for policy-oriented research.

In conclusion, between-visit care constitutes a substantial, persistent and clinically meaningful component of pediatric IBD management. The continuation of this care beyond diagnosis and its close association with disease severity underscore the importance of between-visit care as a core element of care delivery. Recognizing and quantifying this invisible burden is essential to inform accurate workforce planning, equitable reimbursement, and the thoughtful development, deployment, and integration of digital and AI-supported tools and strategies to support both patients and clinicians and sustain high-quality care over the long trajectory of pediatric IBD.

Funding

None declared.

Conflicts of interest

All authors have no conflicts to report.

Ethical considerations

This study was reviewed and approved by the University of California San Diego institutional review board.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.