Research ArticleOriginal Research

Hospital Utilization Among Users of Electronic Nicotine Delivery Systems (ENDS) Compared to Other Tobacco Users

Niharika Khanna, Carissa S. Kwan, Elena N. Klyushnenkova, Colleen Kernan, Michael Dark, Julia Melamed and Janaki Deepak
The Journal of the American Board of Family Medicine September 2026, 39 (1) 167173; DOI: https://doi.org/10.3122/jabfm.2025.250265R1

Abstract

Introduction Use of electronic nicotine delivery systems (ENDS) poses a significant and avoidable risk to respiratory health, yet its effect is not fully understood. This study examines the levels of hospital utilization among those who use ENDS and/or other tobacco and develop respiratory conditions.

Methods 260,149 adult patients were screened for ENDS use between August 2020 and June 2022 at the University of Maryland Medical System. Multivariable nominal logistic regression modeling was used to assess associations between tobacco/nicotine use and emergency department (ED) or inpatient utilization due to respiratory conditions defined by admission or discharge ICD-10 J codes. Covariates included age, sex, race, Elixhauser comorbidity score, and history of COVID-19.

Results Among ENDS-only users, non-Hispanic White and Other-race patients had statistically significant increased healthcare utilization compared to never users (those who never use ENDS or tobacco). 18-44 year-old patients in these two race groups had 1.9-2.3 times higher odds of utilizing the ED compared to those who never used tobacco; and 65+ year-old non-Hispanic White patients had 2.4 times higher odds of inpatient admission compared to patients who never used tobacco. The odds of increased healthcare utilization among those who used both ENDS and other tobacco ranged from 1.4-2.4 and also varied by age and race.

Conclusions ENDS and tobacco use were associated with increased hospital utilization, including ED and inpatient, due to respiratory illnesses in an age- and race-dependent manner.

Introduction

Electronic nicotine delivery systems (ENDS) were initially marketed in the United States (US) to an older demographic as “safe” alternatives to cigarettes.1 ENDS aerosol output contains various toxic substances and heavy metals, and these exposures may predispose individuals to a higher propensity for respiratory infections, including COVID-19.2–5 ENDS use has also been linked to chronic lung disease and may have systemic effects on multiple organs.3

The industry quickly shifted towards marketing ENDS to younger populations with the inclusion of fruit, mint, and other flavor additions.6–10 In 2019, ENDS or vaping product use–associated lung injury (EVALI) was reported.11 Although EVALI has been associated with various different respiratory diseases, it is difficult to capture the full effect of ENDS use on the respiratory system because EVALI is diagnosed through a very narrow set of ICD-10 codes (all J80, all J82, selected J84, and U07.0).12–14

Cook et al suggest that most adults who use ENDS are either current or former cigarette users, making it important to account for cigarette-smoking histories when assessing potential independent effects of ENDS.15–17 Exclusive-, dual-, and poly-tobacco use was associated with higher incidence of bronchitis, pneumonia, or chronic cough.18 Mukerjee et al found that among youths, while ENDS use alone or cigarette use alone demonstrates association between tobacco use and respiratory illness, the association is higher when both ENDS and cigarettes are used together.19 In addition, prior work has found that ENDS users span the age spectrum, fueling the need for studies that examine the differing patterns of ENDS use among various age groups.20,21

The US Preventive Services Task Force recommends that all patients should be asked about their tobacco use.22 Given the front-line role family physicians play in both preventive counseling and acute-care referrals, understanding these associations is vital to informing clinical practice. Having the initial conversation started by a trusted physician can encourage patients to follow through with treatment via a cessation program where dedicated resources are available. There is a need for rigorous scientific data to support family physicians in managing individuals who use ENDS.

The majority of existing ENDS studies focus on EVALI, with limited data available on the effect of ENDS on respiratory health in general. In addition, past studies that examined relationships between tobacco use and healthcare utilization focused on smoking prevalence among already hospitalized patients.23 Our study seeks to address the gap by examining the correlation between ENDS use and associated other tobacco use on a broad range of respiratory conditions, along with a focus on the utilization of the hospital or emergency department (ED). This novel approach utilized health system–level data to better understand ENDS users and their hospital use patterns.

We hypothesized that tobacco users, including ENDS users, of all ages are more likely to become ill from respiratory and other conditions, compared to those who do not use nicotine/tobacco products, and are therefore more likely to use the ED and hospital. The findings aim to inform providers about evolving trends in tobacco-related health risks and provide support for more targeted screening and cessation strategies.

Methods

Ethics Statement

This cross-sectional study was designated as exempt by the University of Maryland Baltimore Institutional Review Board (reference number HP-00093456).

Study Sample Population

The University of Maryland Medical System (UMMS) is a university-based regional healthcare system that provides primary, specialty ambulatory, and inpatient care at 12 hospitals and 150+ ambulatory locations across the state. The UMMS hospitals have 2,458 licensed beds, with 100,985 hospital admissions, 1,230,086 outpatient visits, and 329,547 emergency visits in 2022.24 UMMS is a site for the Cancer Center Cessation Initiative at the Greenebaum Comprehensive Cancer Center where additional tobacco resources are supported through a Moonshot grant from the National Cancer Institute.25 The electronic health record (EHR) links electronically to the Maryland Quitline and to the Tobacco Health Practice.26–28

Our team facilitated the implementation of a comprehensive, EHR-based E-cigarette and Vaping Assessment Tool (EVAT) within the UMMS EHR in August 2020, to allow the collection of accurate ENDS-use data. EVAT was integrated into the hospital system’s EHR to provide detailed documentation for ENDS use (current every day, current some days, former, never).29

Data for this study were retrieved from the UMMS Epic EHR for 260,149 patients, representing a diverse patient population from Maryland (Appendix Figure 1).30 Patients included in the study sample must have been screened using EVAT between August 2020 and June 2022, had data on other tobacco-product use in their EHR social history, and be at least 18 years old. For patients evaluated for tobacco use several times during the study period, the most current status was used.

Variables

EHR data on other tobacco product use, demographics (age, sex, race), and ICD-10 codes for medical diagnoses from January 2019 to September 2022 were subsequently extracted. January 2019 was used as the start date since EVALI had first been noted as a public health emergency in 2019.31

Respiratory diagnosis–related healthcare utilization was the primary outcome, reflecting the maximum level of care the patient received during the observation period. Hospital encounters with at least one respiratory-related illness, defined by an admission or discharge “J” code ICD-10 diagnosis that occurred within the study period, were included in the analysis. Utilization level was assessed for three mutually exclusive categories:

  1. no respiratory diagnosis–related hospital encounter

  2. ED care alone

  3. ever hospitalized (inpatient)

Tobacco product–use status was the primary predictor and was determined based on two variables: ENDS use status (never, current, former) from EVAT and any other tobacco use status (never, current, former) from Epic EHR social history. The primary predictor was categorized into four mutually exclusive groups:

  1. never used any tobacco products

  2. ever only used ENDS (includes current and former users)

  3. ever only used other tobacco products except ENDS (includes current and former users)

  4. ever used ENDS and other tobacco products (includes current and former users)

Covariate measures included comorbid diagnoses, COVID-19 diagnosis, age, sex, and race.

Comorbid diagnoses were synthesized using an Elixhauser comorbidity index score weighted to predict the risk of 30-day, all-cause readmission. The index is a composite score comprised of 38 individual comorbidity measures, with each measure defined through specific ICD-10-CM codes.32 The index score was analyzed as a categorical variable, with scores ≤ 0 placed in the lowest category and all remaining scores divided into quartiles based on the overall sample distribution.

COVID-19 diagnosis was expressed as a binary variable. Any instance of a positive COVID-19 diagnosis during the study period was determined to be a confirmed case. This covariate was included due to our data sample collection occurring during the pandemic to account for its effects.

Age was expressed in years and divided into three categories (18-44, 45-64, and 65+) using guidance from the Population Assessment of Tobacco and Health Study.33 Race was divided into non-Hispanic White, non-Hispanic Black, and Other.34

Statistical Analysis

Descriptive statistics for the study sample were calculated, with frequency and percent reported. Chi-square (χ2) tests for significance of association between respiratory diagnosis–related healthcare utilization and each of the independent variables were analyzed using bivariate analysis. The association between ENDS-use status and respiratory diagnosis–related healthcare utilization was tested using an ordinal multivariable logistic regression model, with and without adjustment of the covariates. Further tests analyzing age, sex, and race as interactive effects were also conducted. Statistical significance for all tests was set at α = 0.05 (two-tailed test) and all analyses were computed using SAS version 9.4 (SAS Institute Inc; Cary, NC).

Results

Population Characteristics

The demographic and clinical characteristics of the study population are shown in Table 1, with a breakdown by tobacco use status shown in Table 2. In the study population, 3,928 (1.5%) ever only used ENDS, 91,986 (35.4%) ever only used tobacco products other than ENDS, and 11,688 (4.5%) were patients who had used both ENDS and other tobacco (Table 1). Among all patients, 37.3% were 18-44 years old, 34.3% were 45-64 years old, and 28.4% were 65+ years old; 59.4% of the patients were women; 61.8% were non-Hispanic White (Table 1). A history of at least one comorbidity contributing to the Elixhauser index was documented in 45.0% of patients and 1.7% had a history of COVID-19 (Table 1).

View this table:
Table 1. Demographic and clinical characteristics for 260,149 UMMS patients evaluated for ENDS use through EVAT between August 2020 and June 2022

Among 260,149 patients in the study population, 20,748 (8.0%) utilized the ED and 27,220 (10.5%) were admitted to the hospital (Table 1). More detailed distribution of the ICD-10 codes associated with admission or discharge are shown in Appendix Table 1. The most frequent ICD-10 codes associated with ED utilization were for chronic lower respiratory diseases followed by acute upper respiratory infections. The most frequent codes associated with inpatient admission were also for chronic lower respiratory diseases, followed by other diseases of the respiratory system.

Among those who went to the ED, 1.9% were ENDS-only users, 45.8% were tobacco-only users, and 6.5% used both (Table 2). However, among those who were admitted to the hospital, the percentage of ENDS-only users and patients who used both ENDS and tobacco decreased to 0.8% and 4.5%, respectively, while the percentage of tobacco-only users increased to 52.9% (Table 2).

View this table:
Table 2. Descriptive Statistics for 260,149 UMMS Adult Patients Evaluated for ENDS Use Through EVAT Between August 2020 and June 2022, by Tobacco-Use Status

Association Between Healthcare Utilization and Tobacco Use

A higher proportion of patients with a history of any type of tobacco use (ENDS only, other tobacco products only, or both) went to the ED compared to never users, with ED visits comprising around 10% of respiratory diagnosis–related hospital encounters for each tobacco-use status group (Table 3). In addition, those who only used other tobacco products had a higher percentage of inpatient admissions (15.6%) compared to patients in each of the other tobacco-use status groups (Table 3). We found in bivariate analyses that tobacco-use status, Elixhauser comorbidity index score, COVID-19 diagnosis, age, sex, and race all have statistically significant associations with healthcare utilization (Table 3).

View this table:
Table 3. Association Between Respiratory Diagnosis–Related Healthcare Utilization, and Demographic and Clinical Characteristics for 260,149 UMMS Adult Patients Evaluated for ENDS Use Through EVAT Between August 2020 and June 2022

In a bivariate nominal logistic regression analysis, ENDS-only users had greater odds of going to the ED (OR = 1.59; 95% CI = 1.43, 1.77; P < 0.0001), but were less likely to be admitted to the hospital (OR = 0.73; 95% CI = 0.63, 0.84; P < 0.0001) compared to patients who never used tobacco (Appendix Table 2). Those who used other tobacco products only had greater odds of either going to the ED (OR = 1.93; 95% CI = 1.88, 1.99; P < 0.0001) or being admitted to the hospital (OR = 2.44; 95% CI = 2.38, 2.51, P < 0.0001) (Appendix Table 2). Those who used both ENDS and other tobacco had even greater odds of going to the ED (OR = 2.05; 95% CI = 1.93, 2.18; P < 0.0001) or being admitted to the hospital (OR = 1.56; 95% CI = 1.46, 1.66; P < 0.0001) (Appendix Table 2).

Association Between Healthcare Utilization and Tobacco Use: Multivariable Logistic Regression Modeling

The effect of tobacco use on healthcare utilization was further assessed in a multivariable nominal logistic regression model adjusting for potential confounding effect of demographic (age, sex, race) and clinical (Elixhauser comorbidity score, history of COVID-19) characteristics. The model also accounted for age and race as effect modifiers for the association between the main predictor and outcome. The effect modification of sex was explored, but was ultimately excluded due to sparseness of data and lack of clinical significance (Appendix Figure 2 and data not shown). The adjusted odds ratios (aOR) comparing tobacco users to never users in age and race strata (controlling for sex, Elixhauser comorbidity score, and COVID-19 diagnoses) are shown in Table 4. aORs for association between healthcare utilization and other demographic and clinical characteristics derived from the same multivariable regression model are shown in Appendix Table 3.

View this table:
Table 4. Association Between Respiratory Diagnosis–Related Healthcare Utilization and Tobacco-Use Status in the Study Population by Age and Race Groups: Results of the Multivariable Nominal Logistic Regression Model

Multivariable nominal logistic regression model for the probability of ED or IP encounter included tobacco-use status, demographic (age, sex, and race), and clinical (Elixhauser comorbidity score and history of COVID-19) characteristics as predictors with “No hospital encounter” as the reference category. Interaction terms for tobacco-use status, age, and race were also included.

Odds ratio estimates for the effect of tobacco-use status on healthcare utilization adjusted for sex, race, Elixhauser comorbidity index, and COVID-19 diagnosis (aOR), 95% CI, and P-values are shown in age and race strata. aOR estimates for covariates other than tobacco-use status (controlling for all other variables in the model) are shown in Appendix Table 3.

Abbreviations: aOR = adjusted odds ratio; CI = confidence interval; ED = emergency department; ENDS = electronic nicotine delivery system; EVAT = e-cigarette and vaping assessment tool; ICD-10 = International Classification of Diseases (10th revision); IP = inpatient admission; NH = non-Hispanic; Pr = probability; UMMS = University of Maryland Medical System.

Emergency Department

ENDS only: Among ENDS-only users, only non-Hispanic White and Other patients aged 18-44 had statistically significant odds of visiting the ED, with a 1.9-fold increase among non-Hispanic Whites (aOR = 1.90; 95% CI = 1.63, 2.22; P < 0.0001) and a 2.3-fold increase among Other patients (aOR = 2.31; 95% CI = 1.63,3.27; P = 0.0048) compared to patients of the same age, race, and other demographic characteristics who never used tobacco (Table 4). Results for Other patients aged 65+ were unable to be generated due to sparseness of data.

Other tobacco only: Associations between exclusive use of tobacco and ED use were statistically significant in all age and race groups. Among 18-44 year-olds, all race groups had approximately a 2-fold increase of ED use compared to those who never used tobacco (non-Hispanic White: aOR = 1.87; 95% CI = 1.73, 2.02; P < 0.0001 | non-Hispanic Black: aOR = 1.98; 95% CI = 1.85, 2.13; P < 0.0001 | Other: aOR = 2.13; 95% CI = 1.79, 2.54, P < 0.0001). Among the 45-64 year-old and the 65+ year-old age groups, the odds of ED use increased approximately 2-fold in non-Hispanic White (45-64: aOR = 2.02; 95% CI = 1.87, 2.13; P < 0.0001 | 65+: aOR = 1.84; 95% CI = 1.71, 1.99; P < 0.0001) and non-Hispanic Black patients (45-64: aOR = 1.97; 95% CI = 1.81, 2.13; P < 0.0001 | 65+: aOR = 1.80; 95% CI = 1.60, 2.03; P < 0.0001), and approximately 1.5-fold among Other patients compared to those who never used tobacco (45-64: aOR = 1.43; 95% CI = 1.13, 1.79; P = 0.0025 | 65+: aOR = 1.52; 95% CI = 1.10, 2.11; P = 0.0117) (Table 4).

Both ENDS and other tobacco users: Among ENDS and tobacco users, the odds of visiting the ED ranged from 1.7 to 2.4 times the odds compared to those who never used tobacco, with 45-64 year-old non-Hispanic White patients having the highest odds (aOR = 2.42; 95% CI = 2.10, 2.79; P < 0.0001) (Table 4).

Inpatient

ENDS only: Among those who only used ENDS, the odds of inpatient admission were only statistically significant among non-Hispanic White patients aged 65+ with an aOR of 2.4 compared to patients who never used tobacco (Table 4).

Other tobacco only: 18-44 year-old patients who use other tobacco products have approximately 1.4 times the odds of inpatient admission compared to those who never used tobacco (non-Hispanic White: aOR = 1.35; 95% CI = 1.22, 1.49; P < 0.0001 | non-Hispanic Black: aOR = 1.45; 95% CI = 1.31, 1.61; P < 0.0001 | Other: aOR = 1.47; 95% CI = 1.19, 1.81; P = 0.0004). The odds varied by race for patients aged 45-64, with non-Hispanic Whites having the highest odds of inpatient admission at 2.1 times when compared to patients of the same age and other demographic characteristics who never used tobacco (aOR = 2.12; 95% CI = 1.99, 2.27; P < 0.0001). Among patients aged 65+, non-Hispanic White and non-Hispanic Black patients both have approximately 1.7 times the odds of inpatient admission compared to patients who never used tobacco (non-Hispanic White: aOR = 1.73; 95% CI = 1.64, 1.81; P < 0.0001 | non-Hispanic Black: aOR = 1.67; 95% CI = 1.52, 1.83; P < 0.0001) while the odds for patients of the Other race group were not statistically significant (Table 4).

Both ENDS and other tobacco users: Among patients who used tobacco products other than ENDS, the odds of inpatient admission were statistically significant among all age groups for non-Hispanic White patients (18-44: aOR = 1.36; 95% CI = 1.19, 1.55; P < 0.0001 | 45-64: aOR = 2.28; 95% CI = 2.01, 2.31; P < 0.0001 | 65+: aOR = 2.28; 95% CI = 1.92, 2.69; P < 0.0001) and among those aged 45+ for non-Hispanic Black patients (45-64: aOR = 1.78; 95% CI = 1.38, 2.31; P < 0.0001 | 65+: aOR = 2.16; 95% CI = 1.39, 3.35; P = 0.0006) compared to those who never used tobacco (Table 4).

Discussion

The marketing of ENDS in the US as cessation devices set the stage for public consideration of their benefits despite the lack of scientific evidence to suggest their safety. ENDS are newer ways of nicotine delivery utilized by various demographics with limited understanding of their harms. This study explored the hypothesis that lung effects from exposure to ENDS aerosols may increase the need for patients with respiratory illnesses to utilize the ED and hospital.

Any tobacco use, including ENDS, has a greater likelihood of requiring respiratory illness–related ED or inpatient hospital utilization. Specifically, 14.7% of patients in the study sample who used any type of tobacco (ENDS or traditional) were hospitalized for a respiratory-related illness. This is consistent with current literature, which found that around one in six (16%) of all hospital admissions for respiratory diseases in 2022-2023 were estimated to be related to smoking.35 The results of our study also expanded upon previous research by exploring how this association varied between different age and race groups.

Patients who only used other tobacco products (excluding ENDS) had increased odds of all levels of healthcare utilization when compared to never users even after adjusting for covariates and stratifying by age and race. These results suggest that tobacco use is a risk factor for severe respiratory disease and increased healthcare utilization for all populations.36 However, we found that this effect varied by age and race with regards to ENDS-only users as well as patients who used both ENDS and other tobacco products. The findings from this study cohort initially seem to suggest a potentially slight protective effect for inpatient admissions among ENDS-only users. However, this phenomenon disappeared or became statistically insignificant after stratifying by age and race. Furthermore, additional analysis of the regression model indicated that non-Hispanic White patients who used both ENDS and other tobacco products had statistically significant increased healthcare utilization compared to those who only used other tobacco products, suggesting that use of combustible tobacco and ENDS together may have higher likelihood of severe disease that may result in hospital use (data not shown).

The hypothesis for this study was corroborated: ED and inpatient utilization increased among tobacco users, including ENDS users, compared to those who did not use tobacco. In addition, our study indicates that ENDS use, especially in combination with other tobacco products, can lead to increased risk of healthcare utilization, providing evidence that ENDS may not be an ideal tobacco-cessation method. EVAT was developed at UMMS not only to improve ENDS-use data collection, but also to encourage primary care providers to discuss cessation options with their patients.29 Education on best practices for using the EVAT in Epic EHR, and information on demographic-usage trends obtained from this study, will provide guidance on how to communicate targeted patient options for evidence-based cessation therapies.

Working with a population-based, retrospective EHR data sample has allowed identification of patients who were not previously flagged with ENDS use–related diagnoses. In addition, our work is one of the few observational studies in the senior-aged 65+ population, and one of the first to initiate the discussion on the importance of querying and recognizing the use of ENDS devices, in addition to the use of other tobacco, in every patient seen at the hospital or ambulatory setting. Tobacco/nicotine treatment may play a role in reducing the likelihood of getting sick and utilizing the healthcare system, and family physicians can play a key role in initiating conversations and identifying ENDS use in the ambulatory setting.

There are some limitations of tobacco-use history documentation and data gathering from the EHR. Historical data in the tobacco history section get lost because of EHR system settings that allow the overwriting of history at each visit. The data presented in the EHR consist of the observations and documentation from the patient’s most recent visit. Additionally, tobacco-use data may not be refreshed in certain low-acuity visits during which a complete past history is not taken by the clinician. As a result, complete historical tobacco/nicotine product-use data is not available. The data acquired is a snapshot in time and does not clearly define the duration and intensity of use. Our data does not allow us to identify the start dates for ENDS use, nor does it provide information about the reasons for initiating the use of multiple forms of tobacco/nicotine.

Another limitation of our study is the heterogeneity of the patient group that was categorized as users of both ENDS and other tobacco products. Because data on start dates and duration of use were unavailable, patients who were former ENDS users that switched to using other tobacco products only, and vice versa, were unidentifiable from each other. Additionally, this category also includes patients who concurrently use conventional combustible tobacco and ENDS. Concurrent dual users themselves are also a heterogenous population group comprised of individuals who started using ENDS in young age and subsequently progressed to other types of tobacco use, in addition to older individuals who may have started smoking at an early age and are trying emerging nicotine products, possibly as a means to cut smoking tobacco.37 Although the odds of healthcare utilization among users who used both ENDS and other tobacco were not as high compared to those who only used other tobacco, our data show that use of both conventional combustible tobacco and ENDS is detrimental to a patient’s health.

This cohort includes data from the UMMS catchment area, and while demographic diversity was present, caution should be taken when extending the resulting observations to other parts of the US, given that tobacco usage, demographic profiles, and healthcare access can differ from region to region. Nonetheless, the observed associations are relevant to other healthcare system settings within the US.

Additional studies are needed to elucidate a complete tobacco/nicotine history, including duration and intensity of ENDS and other tobacco use. The current analyses lacked socioeconomic status (SES) variables, which is a known confounder in both tobacco use and healthcare access.38–40 Future models should incorporate SES and related social determinants of health to further refine these findings. The importance of including all ages of individuals, especially children and the elderly, is clear from our data, and suggests the need to examine the impact on those who have not been included in prior research. Further breakdown of the different subgroups of dual users and their tobacco/nicotine-use patterns may also be beneficial towards developing targeted tobacco-treatment plans for individuals who use tobacco/nicotine. Lastly, further analyses of tobacco/nicotine use on specific respiratory conditions, such as ambulatory care–sensitive conditions versus more severe acute conditions, would help support revisions to ambulatory care policies to promote new workflows and vital-sign monitoring for all tobacco users; this could facilitate early therapeutic interventions and minimize preventable increased healthcare utilization.

It is important to educate all patients, including youth, adults, and seniors, on the detrimental effects of ENDS use. It is also important to train youth- and adult-serving family physicians, teachers, nurses, and parents on how to effectively talk about ENDS. This will ultimately lead to decreases in initiation of ENDS use and encourage individuals to seek help with cessation.

Conflicts of Interest

None.

Corresponding Author

Niharika Khanna, MD, MBBS, DGO, Professor, Family and Community Medicine, Associate Chair, Population Health Sciences, University of Maryland School of Medicine, nkhanna{at}som.umaryland.edu

This article was externally peer reviewed.

Data Availability

Data is not publicly available due to HIPAA regulations.

Appendices

Appendix Figure 1.
Appendix Figure 1. Flow of UMMS Patients Included in Study Sample

Notes: Flow chart depicting inclusion criteria for the study. To be included in the analysis, a patient must have been (a) evaluated for ENDS use using EVAT between August 2020 and June 2022, (b) had data on other tobacco product use in the EHR social history, and (c) be at least 18 years old. The final sample included 260,149 adult patients. For patients evaluated for tobacco use several times during the study period, the most current status was used.

Abbreviations: EHR = electronic health record; ENDS = electronic nicotine delivery system; EVAT = e-cigarette and vaping assessment tool; UMMS = University of Maryland Medical System.

View this table:
Appendix Table 1. Respiratory Illnesses Defined by Either Admission or Discharge “J” ICD-10 Codes in the Study Population
View this table:
Appendix Table 2. Association Between Respiratory Diagnosis–Related Healthcare Utilization and Tobacco-Use Status in the Study Population: Results of the Bivariate Nominal Logistic Regression Analysis
A graph of different types of lines AI-generated content may be incorrect.
Appendix Figure 2. Prevalence of Respiratory Diagnosis–Related Healthcare Utilization in Demographic Groups by Tobacco-Use Status for 260,149 UMMS Adult Patients Evaluated for ENDS Use Through EVAT Between August 2020 and June 2022

Notes: Data for 260,149 adult UMMS patients evaluated for ENDS use through EVAT between August 2020 and June 2022 are shown. Respiratory diagnosis–related healthcare utilization was defined by presence of the admission or discharge “J” ICD-10 code and contained three mutually exclusive categories: None (absence of admission or discharge “J” ICD-10 code), ED, and IP. Tobacco-use status was defined as never used, ENDS only (current and former), other tobacco products only (current and former; mostly cigarettes, but may include cigars, pipe, or other smoking devices except ENDS), and users who have used both ENDS and any other form of tobacco.

Prevalence of ED (A, B, C) and IP (D, E, F) is shown in age (A, D), sex (B, E), and race (C, F) demographic groups by tobacco-use status. Age is depicted as follows: 18-44, circles/solid line; 45-64, squares/dotted line; 65+, triangles/dashed line. Sex is depicted as follows: Female, circles/dashed line; Male, squares/solid line. Race is depicted as follows: NH White, circles/dotted line; NH Black, squares/dashed line; Other, triangles/solid line.

Abbreviations: ED = emergency department; ENDS = electronic nicotine delivery system; EVAT = e-cigarette and vaping assessment tool; ICD-10 = International Classification of Diseases (10th revision); IP = inpatient admission; NH = non-Hispanic; UMMS = University of Maryland Medical System.

View this table:
Appendix Table 3. Association Between Respiratory Diagnosis–Related Healthcare Utilization, and Demographic and Clinical Characteristics in the Study Population: Full Results of the Multivariable Nominal Logistic Regression Model.
  • Received for publication July 11, 2025.
  • Accepted for publication November 10, 2025.

References

  1. ↵
    1. Fadus M. C.,
    2. Smith T. T.,
    3. Squeglia L. M.
    (2019) The rise of e-cigarettes, pod mod devices, and JUUL among youth: factors influencing use, health implications, and downstream effects. Drug Alcohol Depend 201:85–93, doi:10.1016/j.drugalcdep.2019.04.011, https://doi.org/10.1016/j.drugalcdep.2019.04.011. .
    OpenUrlCrossRefPubMed
  2. ↵
    1. Leigh N. J.,
    2. Lawton R. I.,
    3. Hershberger P. A.,
    4. Goniewicz M. L.
    (2016) Flavourings significantly affect inhalation toxicity of aerosol generated from electronic nicotine delivery systems (ENDS). Tob Control 25(Suppl 2):ii81–ii87, doi:10.1136/tobaccocontrol-2016-053205, https://doi.org/10.1136/tobaccocontrol-2016-053205. .
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Park J. A.,
    2. Crotty Alexander L. E.,
    3. Christiani D. C.
    (2022) Vaping and lung inflammation and injury. Annu Rev Physiol 84(1):611–629, doi:10.1146/annurev-physiol-061121-040014, https://doi.org/10.1146/annurev-physiol-061121-040014. .
    OpenUrlCrossRefPubMed
  4. ↵
    1. Feeney S.,
    2. Rossetti V.,
    3. Terrien J.
    (2022) E-cigarettes—a review of the evidence—harm versus harm reduction. Tob Use Insights 15, doi:10.1177/1179173X221087524, https://doi.org/10.1177/1179173X221087524. .
    OpenUrlCrossRefPubMed
  5. ↵
    1. Miyashita L.,
    2. Foley G.
    (2020) E-cigarettes and respiratory health: the latest evidence. J Physiol 598(22):5027–5038, doi:10.1113/JP279526, https://doi.org/10.1113/JP279526. .
    OpenUrlCrossRefPubMed
  6. ↵
    1. Bhatt J. M.,
    2. Ramphul M.,
    3. Bush A.
    (2020) An update on controversies in e-cigarettes. Paediatr Respir Rev 36:75–86, doi:10.1016/j.prrv.2020.09.003, https://doi.org/10.1016/j.prrv.2020.09.003. .
    OpenUrlCrossRefPubMed
  7. ↵
    1. Loukas A.,
    2. Paddock E. M.,
    3. Li X.,
    4. Harrell M. B.,
    5. Pasch K. E.,
    6. Perry C. L.
    (2019) Electronic nicotine delivery systems marketing and initiation among youth and young adults. Pediatrics 144(3):e20183601, doi:10.1542/peds.2018-3601, https://doi.org/10.1542/peds.2018-3601. .
    OpenUrlCrossRefPubMed
  8. ↵
    1. Cruz T. B.,
    2. McConnell R.,
    3. Low B. W..,
    4. et al.
    (2019) Tobacco marketing and subsequent use of cigarettes, e-cigarettes, and hookah in adolescents. Nicotine Tob Res 21(7):926–932, doi:10.1093/ntr/nty107, https://doi.org/10.1093/ntr/nty107. .
    OpenUrlCrossRefPubMed
  9. ↵
    1. Collins L.,
    2. Glasser A. M.,
    3. Abudayyeh H.,
    4. Pearson J. L.,
    5. Villanti A. C.
    (2019) E-cigarette marketing and communication: how e-cigarette companies market e-cigarettes and the public engages with e-cigarette information. Nicotine Tob Res 21(1):14–24, doi:10.1093/ntr/ntx284, https://doi.org/10.1093/ntr/ntx284. .
    OpenUrlCrossRefPubMed
  10. ↵
    1. Harrington K. F.,
    2. Hull N. C.,
    3. Akindoju O..,
    4. et al.
    (2014) Electronic cigarette awareness, use history, and expected future use among hospitalized cigarette smokers. Nicotine Tob Res 16(11):1512–1517, doi:10.1093/ntr/ntu054, https://doi.org/10.1093/ntr/ntu054. .
    OpenUrlCrossRefPubMed
  11. ↵
    1. James A. E.,
    2. Sutphin B.,
    3. Fleming P.,
    4. McCormick D.,
    5. Balamurugan A.
    (2020) Characteristics of e-cigarette, or vaping, product use-associated lung injury (EVALI) patients. J Ark Med Soc 117(5):110–112, pmid:37179818. .
    OpenUrlPubMed
  12. ↵
    1. Zulfiqar H.,
    2. Sankari A.,
    3. Rahman O.
    (2025) StatPearls (StatPearls Publishing), https://www.ncbi.nlm.nih.gov/books/NBK560656/. . Vaping-Associated Pulmonary Injury. Updated June 25, 2023. 2025-7-10.
  13. ↵
    1. Layden J. E.,
    2. Ghinai I.,
    3. Pray I..,
    4. et al.
    (2020) Pulmonary illness related to e-cigarette use in Illinois and Wisconsin—final report. N Engl J Med 382(10):903–916, doi:10.1056/NEJMoa1911614, https://doi.org/10.1056/NEJMoa1911614. .
    OpenUrlCrossRefPubMed
  14. ↵
    1. Centers for Disease Control and Prevention
    (10 17, 2019) ICD-10-CM Official Coding Guidelines—Supplement: Coding encounters related to E-cigarette, or Vaping, Product Use. . 2025-7-10. https://www.cdc.gov/nchs/data/icd/Vapingcodingguidance2019_10_17_2019.pdf.
  15. ↵
    1. Cook S. F.,
    2. Hirschtick J. L.,
    3. Fleischer N. L..,
    4. et al.
    (2023) Cigarettes, ENDS use, and chronic obstructive pulmonary disease incidence: a prospective longitudinal study. Am J Prev Med 65(2):173–181, doi:10.1016/j.amepre.2023.01.038, https://doi.org/10.1016/j.amepre.2023.01.038. .
    OpenUrlCrossRefPubMed
  16. ↵
    1. Bandi P.,
    2. Cahn Z.,
    3. Sauer A.G..,
    4. et al.
    (2021) Trends in e-cigarette use by age group and combustible cigarette smoking histories, U.S. adults, 2014–2018. Am J Prev Med 60(2):151–158, doi:10.1016/j.amepre.2020.07.026, https://doi.org/10.1016/j.amepre.2020.07.026. .
    OpenUrlCrossRefPubMed
  17. ↵
    1. Mayer M.,
    2. Reyes-Guzman C.,
    3. Grana R.,
    4. Choi K.,
    5. Freedman N.D.
    (2020) Demographic characteristics, cigarette smoking, and e-cigarette use among US adults. JAMA Netw Open 3(10):e2020694, doi:10.1001/jamanetworkopen.2020.20694, https://doi.org/10.1001/jamanetworkopen.2020.20694. .
    OpenUrlCrossRefPubMed
  18. ↵
    1. Zavala-Arciniega L.,
    2. Cook S.,
    3. Hirschtick J. L..,
    4. et al.
    (2024) Longitudinal associations between exclusive, dual and polytobacco use and respiratory illness among youth. BMC Public Health 24(1):2159, doi:10.1186/s12889-024-19582-8, https://doi.org/10.1186/s12889-024-19582-8. .
    OpenUrlCrossRefPubMed
  19. ↵
    1. Mukerjee R.,
    2. Hirschtick J. L.,
    3. Arciniega L. Z..,
    4. et al.
    (2024) ENDS, cigarettes, and respiratory illness: longitudinal associations among U.S. youth. Am J Prev Med 66(5):789–796, doi:10.1016/j.amepre.2023.12.005, https://doi.org/10.1016/j.amepre.2023.12.005. .
    OpenUrlCrossRefPubMed
  20. ↵
    1. Zare S.,
    2. Nemati M.,
    3. Zheng Y.
    (2018) A systematic review of consumer preference for e-cigarette attributes: flavor, nicotine strength, and type. PLOS One 13(3):e0194145, doi:10.1371/journal.pone.0194145, https://doi.org/10.1371/journal.pone.0194145. .
    OpenUrlCrossRefPubMed
  21. ↵
    1. Gades M. S.,
    2. Alcheva A.,
    3. Riegelman A. L.,
    4. Hatsukami D. K.
    (2022) The role of nicotine and flavor in the abuse potential and appeal of electronic cigarettes for adult current and former cigarette and electronic cigarette users: a systematic review. Nicotine Tob Res 24(9):1332–1343, doi:10.1093/ntr/ntac073, https://doi.org/10.1093/ntr/ntac073. .
    OpenUrlCrossRefPubMed
  22. ↵
    1. US Preventive Services Task Force,
    2. Krist A. H.,
    3. Davidson K. W..,
    4. et al.
    (2021) Interventions for tobacco smoking cessation in adults, including pregnant persons: US Preventive Services Task Force recommendation statement. JAMA 325(3):265, doi:10.1001/jama.2020.25019, https://doi.org/10.1001/jama.2020.25019. .
    OpenUrlCrossRefPubMed
  23. ↵
    1. Regan S.,
    2. Viana J. C.,
    3. Reyen M.,
    4. Rigotti N. A.
    (2012) Prevalence and predictors of smoking by inpatients during a hospital stay. Arch Intern Med 172(21):1670–1674, doi:10.1001/2013.jamainternmed.300, https://doi.org/10.1001/2013.jamainternmed.300. .
    OpenUrlCrossRefPubMed
  24. ↵
  25. ↵
    (3 24, 2025) University of Maryland Marlene and Stewart Greenebaum Comprehensive Cancer Center. . 2025-7-10. https://www.cancer.gov/research/infrastructure/cancer-centers/find/umdgreenebaum.
  26. ↵
    1. Piasecki T. M.,
    2. Smith S. S.,
    3. Baker T. B..,
    4. et al.
    (2023) Smoking status, nicotine medication, vaccination, and COVID-19 hospital outcomes: findings from the COVID EHR cohort at the University of Wisconsin (CEC-UW) study. Nicotine Tob Res 25(6):1184–1193, doi:10.1093/ntr/ntac201, https://doi.org/10.1093/ntr/ntac201. .
    OpenUrlCrossRefPubMed
  27. ↵
    1. Khanna N.,
    2. Klyushnenkova E.,
    3. Rao V.,
    4. Siegel N.,
    5. Wolfe S.
    (2021) Electronic referrals to the tobacco Quitline: implementation strategies in a large health system to optimize delivery of tobacco cessation to patients. Transl Behav Med 11(5):1107–1114, doi:10.1093/tbm/ibaa094, https://doi.org/10.1093/tbm/ibaa094. .
    OpenUrlCrossRefPubMed
  28. ↵
    1. Khanna N.,
    2. Klyushnenkova E. N.,
    3. Quinn D.,
    4. Wolfe S.
    (2023) Patient engagement by the tobacco Quitline after electronic referrals. Nicotine Tob Res 25(1):94–101, doi:10.1093/ntr/ntac190, https://doi.org/10.1093/ntr/ntac190. .
    OpenUrlCrossRefPubMed
  29. ↵
    1. Khanna N.,
    2. Klyushnenkova E.,
    3. Gaynor A..,
    4. et al.
    (2023) Integrating a systematic, comprehensive e-cigarette and vaping assessment tool into the electronic health record. J Am Board Fam Med 36(3):405–413, doi:10.3122/jabfm.2022.220410R1, https://doi.org/10.3122/jabfm.2022.220410R1. .
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. United States Census Bureau
    Maryland. . 2025-7-11. https://data.census.gov/profile/Maryland?g=040XX00US24.
  31. ↵
    1. Centers for Disease Control and Prevention
    (8 3, 2021) Outbreak of Lung Injury Associated with the Use of E-Cigarette, or Vaping, Products. . 2025-7-11. https://archive.cdc.gov/www_cdc_gov/tobacco/basic_information/e-cigarettes/severe-lung-disease.html.
  32. ↵
    1. Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project (HCUP)
    (10, 2021) User Guide: Elixhauser Comorbidity Software Refined for ICD-10-CM Diagnoses, v2022.1 https://hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/CMR-User-Guide-v2022-1.pdf. . 2025-7-11.
  33. ↵
    1. National Institutes of Health
    PATH: Population Assessment of Tobacco and Health. . 2025-7-11. https://pathstudy.nih.gov/.
  34. ↵
    1. Khanna N.,
    2. Klyushnenkova E. N.,
    3. Kaysin A.
    (2021) Association of COVID-19 with race and socioeconomic factors in family medicine. J Am Board Fam Med 34(Supplement):S40–S47, doi:10.3122/jabfm.2021.S1.200338, https://doi.org/10.3122/jabfm.2021.S1.200338. .
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. NHS England
    (12 16, 2023) Hospital admissions due to smoking up nearly 5% last year, NHS data shows. . 2025-7-11. https://www.england.nhs.uk/2023/12/hospital-admissions-due-to-smoking-up-nearly-5-per-cent-last-year-nhs-data-shows/.
  36. ↵
    1. Centers for Disease Control and Prevention
    (5 15, 2024) Benefits of Quitting Smoking. . 2025-7-11. https://www.cdc.gov/tobacco/about/benefits-of-quitting.html.
  37. ↵
    1. Eaton D. L.,
    2. Kwan L. Y.,
    3. Stratton K.
    , eds (2018) Public Health Consequences of E-Cigarettes (National Academies Press (US)), https://www.ncbi.nlm.nih.gov/books/NBK507169/. . Combustible tobacco cigarette smoking among youth and young adults. 2025-7-11.
  38. ↵
    1. Centers for Disease Control and Prevention
    (5 15, 2024) People With Low Socioeconomic Status Experience a Health Burden From Commercial Tobacco. . 2025-10-10. https://www.cdc.gov/tobacco-health-equity/collection/low-ses-health-burden.html.
  39. ↵
    1. Garrett B. E.,
    2. Martell B. N.,
    3. Caraballo R. S.,
    4. King B. A.
    (2019) Socioeconomic Differences in cigarette smoking among sociodemographic groups. Prev Chronic Dis 16:180553, doi:10.5888/pcd16.180553, https://doi.org/10.5888/pcd16.180553. .
    OpenUrlCrossRef
  40. ↵
    1. McMaughan D. J.,
    2. Oloruntoba O.,
    3. Smith M. L.
    (2020) Socioeconomic status and access to healthcare: interrelated drivers for healthy aging. Front Public Health 8:231, doi:10.3389/fpubh.2020.00231, https://doi.org/10.3389/fpubh.2020.00231. .
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Keywords