Abstract
Continuous pulse oximetry (cSpO2) monitoring for hospitalized children with bronchiolitis not requiring supplemental oxygen is an established form of medical overuse. We examined whether clinician team structure influences cSpO2 overuse. In this secondary analysis of the Eliminating Monitor Overuse (EMO) Trial (NCT05132322), we compared overuse by frontline clinician type (resident physician versus attending physician or advanced practice provider [APP]) during the active cSpO2 deimplementation and sustainment phases using multivariable mixed-effects logistic regression. Among 4,167 inpatient observations in 36 hospitals, 61% had their frontline care provided by residents, 27% by attendings, and 10% by APPs. In adjusted analyses, the odds of overuse were 47% higher among patients with frontline care delivered by attendings compared to residents (aOR 1.47 [95% CI 1.12–1.93], p=0.01), with no difference observed for APPs compared to residents. These findings suggest that cSpO2 overuse may be attenuated when residents deliver frontline patient care rather than attendings alone.
INTRODUCTION
Continuous pulse oximetry (cSpO2) use in children with bronchiolitis who are not receiving supplemental oxygen is discouraged by national guidelines and widely recognized as medical overuse,1 contributing to increased alarm burden and healthcare utilization without improving patient outcomes.2 Prior studies have elucidated factors associated with cSpO2 overuse including age and prematurity, overnight monitoring,3 and low bronchiolitis volume hospital units.4 However, it is unknown whether the clinician team structure – specifically, the type of frontline clinician (i.e., responsible for placing orders and serving as the first contact for nurses) – influences cSpO2 overuse. Our objective was to compare cSpO2 overuse by team structure (sub-intern, intern or resident physicians [“residents”]; advanced practice providers [APPs]; or attending physicians in frontline clinician role). We hypothesized that teams with residents serving as the frontline clinician would have higher rates of cSpO2 overuse than teams with frontline care delivered by attendings or APPs, potentially reflecting greater safety concerns among less experienced trainees.
METHODS
Study Design and Setting
We performed a substudy of the Eliminating Monitor Overuse (EMO) SpO2 Trial, a cluster-randomized hybrid type 3 effectiveness-deimplementation trial conducted in the Pediatric Research in Inpatient Settings (PRIS) Network. This trial compared the effects of an unlearning-only strategy (educational outreach and audit & feedback) versus an unlearning plus substitution strategy (EHR clinical decision support) on sustainment of cSpO2 deimplementation in children with bronchiolitis in room air; the study protocol and statistical analysis plan were previously published.5,6 The three phases of the trial were baseline measurement (Phase 1, P1), active deimplementation (P2), and sustainment (P3, after withdrawal of the unlearning-based strategies). Data for this substudy were collected during P2 (1/20/2023 – 5/20/2023) and P3 (12/1–2023 – 3/31/2024).
Patient Population
The parent trial included infants with a primary diagnosis of bronchiolitis aged 8 weeks through 23 months who are non-intensive care, non-emergency department, non-step-down inpatient units cared for by a generalized inpatient service. Infants with major comorbidities, COVID-19, and gestational age < 28 weeks were excluded.5
Data Collection
In the EMO SpO2 Trial, continuous pulse oximetry monitoring status was ascertained through cross-sectional observational data collection rounds conducted by staff at each hospital. In addition to chart review, this included visualizing the cSpO2 waveform at the bedside or using secure remote viewing platforms. Specific to this substudy, during the P2 and P3 phases, we incorporated into the REDCap data collection form a prompt asking staff to complete the following optional question: “Which type of provider is the front-line clinician caring for this patient (i.e., responsible for placing orders, first contact for nurses)?” with response options of sub-intern, intern or resident physician (“residents”), fellow, advanced practice providers (APP, e.g., NP or PA), attending physicians, or other.
Data Analysis
We summarized study variables overall and stratified by clinician team structure by count and percentage. Study variables included (a) patient or observation-level covariates used in prior analyses in the EMO SpO2 portfolio (age combined with preterm birth, time since weaning from supplemental oxygen or flow, documented history of apnea or cyanosis during the present illness, prior intensive care unit stay during present hospitalization, presence of an enteral feeding tube, history of neurologic impairment, and nighttime observation);3,4 (b) study phase (P2 or P3); (c) intervention arm (education + audit-and-feedback versus education + audit-and-feedback + clinical decision support [CDS]); and (d) additional pre-specified hospital-level covariates: presence of pre-existing CDS for cSpO2 use prior to the start of the trial, and hospital type (freestanding or not freestanding). We also examined clinician team structure at the hospital level by describing the percentage of observations corresponding to each team structure within each hospital.
In the primary analysis comparing cSpO2 overuse by clinician team structure, we fit univariable (unadjusted) and multivariable (adjusted) hierarchical mixed-effects logistic regression models, with hospital- and patient-specific random intercepts to account for clustering of observations within patients and hospitals. The covariates adjusted in the multivariable modeling included patient- and hospital-level variables. Cross-tabulation was used to evaluate the extent of overlap and sparse cell counts among the binary hospital-level variables (intervention arm, hospital type, presence of pre-existing CDS) as an indicator of potential collinearity (Appendix Table 1). To evaluate robustness, we conducted sensitivity analyses adjusting for (a) intervention arm alone, (b) intervention arm and hospital type, and (c) intervention arm and pre-existing CDS, and (d) all 3 hospital-level variables. We used predictive margins to estimate the adjusted overuse rates for each clinician team structure. All analyses were performed using Stata version 19.
RESULTS
We identified 4,167 observations of infants with bronchiolitis across 36 hospitals. Clinician team structure was recorded for 4,134 observations (99.2%). Of these, 2,550 (61%) involved frontline care provided by residents, 1,145 (27%) by attendings, 427 (10%) by APPs, 9 (0.2%) by fellows, and 3 (0.1%) by other clinician type (Table 1). After excluding 83 observations with missing covariates, 4,051 observations were included in the adjusted regression analysis. The distribution of clinician team structure by hospital is shown in the Appendix Figure.
Table 1.
Characteristics of Study Population and Continuous Pulse Oximetry Overuse
| Patient Observations, n (column %)a,b | ||||
|---|---|---|---|---|
| Intern/Resident (n=2,550) |
Attending Physician (n=1,145) |
Advanced Practice Provider (n=427) |
Total (N=4,134)a |
|
| Study Phase | ||||
| Phase 2 (Active deimplementation) | 1,299 (50.9) | 544 (47.5) | 180 (42.2) | 2,034 (49.2) |
| Phase 3 (Sustainment)c | 1,251 (49.1) | 601 (52.5) | 247 (57.8) | 2,100 (50.8) |
| Combined Age and Preterm Status | ||||
| 8weeks-5mo & Not Preterm | 653 (25.6) | 330 (28.9) | 109 (25.5) | 1,097 (26.5) |
| 8weeks-5mo & Preterm | 53 (2.0) | 27 (2.4) | 9 (2.1) | 89 (2.2) |
| 6–11mo & Not Preterm | 903 (35.4) | 412 (36.0) | 160 (37.5) | 1,477 (35.7) |
| 6–11mo & Preterm | 76 (3.0) | 29 (2.5) | 15 (3.5) | 120 (2.9) |
| 12–17mo & Not Preterm | 481 (18.9) | 204 (17.8) | 76 (17.8) | 764 (18.5) |
| 12–17mo & Preterm | 32 (1.3) | 17 (1.5) | 2 (0.5) | 51 (1.2) |
| 18–24mo & Not Preterm | 310 (12.2) | 116 (10.1) | 55 (12.9) | 483 (11.7) |
| 18–24mo & Preterm | 27 (1.1) | 7 (0.6) | 1 (0.2) | 35 (0.8) |
| Time Off Supplemental Oxygen or Flow | ||||
| Never on O2 | 471 (18.5) | 262 (22.9) | 127 (29.7) | 862 (20.9) |
| Off 1-<2 hours | 221 (8.7) | 52 (4.5) | 25 (5.9) | 300 (7.3) |
| Off 2-<4 hours | 376 (14.7) | 156 (13.6) | 48 (11.2) | 580 (14.0) |
| Off 4-<6 hours | 230 (9.0) | 122 (10.7) | 38 (8.9) | 395 (9.6) |
| Off 6-<12 hours | 374 (14.7) | 175 (15.3) | 71 (16.6) | 621 (15.0) |
| Off 12-<24 hours | 495 (19.4) | 225 (19.7) | 73 (17.1) | 794 (19.2) |
| Off >=24 hours | 377 (14.8) | 153 (13.4) | 45 (10.5) | 576 (13.9) |
| Neurological Impairment | ||||
| No | 2,474 (97.0) | 1,115 (97.4) | 421 (98.6) | 4,022 (97.3) |
| Yes | 37 (1.5) | 12 (1.0) | 4 (0.9) | 53 (1.3) |
| Prior ICU Stay During Present Hospitalization | ||||
| No | 2,020 (79.2) | 972 (84.9) | 375 (87.8) | 3,378 (81.7) |
| Yes | 492 (19.3) | 155 (13.5) | 50 (11.7) | 698 (16.9) |
| Presence of Enteral Feeding Tube | ||||
| No | 2,389 (93.7) | 1,093 (95.5) | 399 (93.4) | 3,893 (94.2) |
| Yes | 161 (6.3) | 52 (4.5) | 28 (6.6) | 241 (5.8) |
| Overnight Observation | ||||
| No | 2,097 (82.2) | 893 (78.0) | 347 (81.3) | 3,344 (80.9) |
| Yes | 453 (17.8) | 252 (22.0) | 80 (18.7) | 790 (19.1) |
| Hospital Type | ||||
| Not Freestanding (N=10 hospitals) | 429 (16.8) | 247 (21.6) | 56 (13.1) | 735 (17.8) |
| Freestanding (N=26 hospitals) | 2,121 (83.2) | 898 (78.4) | 371 (86.9) | 3,399 (82.2) |
| Presence of Pre-Existing CDS | ||||
| No (N=9 hospitals) | 533 (20.9) | 68 (5.9) | 9 (2.1) | 611 (14.8) |
| Yes (N=27 hospitals) | 2,017 (79.1) | 1,077 (94.1) | 418 (97.9) | 3,523 (85.2) |
| Intervention Arm | ||||
| Education + A&F (N=18 hospitals) | 1,489 (58.4) | 450 (39.3) | 319 (74.7) | 2,264 (54.8) |
| Education + A&F + CDS (N=18 hospitals) | 1,061 (41.6) | 695 (60.7) | 108 (25.3) | 1,870 (45.2) |
| Continuously Monitored with Pulse Oximetry, Observed | ||||
| No | 1,780 (69.8) | 814 (71.1) | 331 (77.5) | 2,930 (70.9) |
| Yes | 770 (30.2) | 331 (28.9) | 96 (22.5) | 1,204 (29.1) |
Abbreviations: O2, oxygen; ICU, intensive care unit; A&F, audit-and-feedback; CDS, clinical decision support
A total of 4,167 observations were recorded in the study. Clinician team structure was documented in 4,134 observations. Observations in which the frontline clinician was classified as fellow (n=9), other (n=3), or missing (n=33) are omitted from the table due to low sample sizes.
Column percentages may not sum to 100% due to missing data
Following withdrawal of unlearning-based implementation strategies (audit-and-feedback and educational outreach)
cSpO2 overuse occurred in 1,204 observations (29.1%). In unadjusted mixed-effects logistic analyses, odds of overuse were 31% higher among patients with frontline care led by attendings compared to residents (OR = 1.31 [95% CI 1.03–1.67], p=0.03). In multivariable mixed-effects modeling after adjustment for covariates, odds of overuse were 47% higher among patients with frontline care led by attendings compared to residents (aOR 1.47 [95% CI 1.12–1.93], p=0.01). This corresponds to a 5.2-percentage point difference (95% CI 1.6–8.9) in adjusted overuse rates [Table 2]. Results were nearly identical across models adjusting for distinct combinations of hospital-level covariates (Appendix Table 2). There was no difference in cSpO2 overuse among patients with frontline care led by APPs compared to residents (aOR 0.91, 95% CI 0.61–1.35, p=0.65).
Table 2.
Continuous Pulse Oximetry Monitoring by Clinician Team Structure
| Frontline Clinician Role | Number of Observations (N = 4,051)a | Unadjusted Overuse, % of Observations (95% CI)b | Adjusted Overuse, % of Observations (95% CI)c |
|---|---|---|---|
| Intern/Resident | 2,490 | 32.1 (26.7 – 37.4) | 30.7 (26.0 – 35.5) |
| Attending | 1,124 | 36.5 (30.3 – 42.7) | 36.0 (30.4 – 41.5) |
| Advanced Practice Provider | 425 | 30.9 (23.7 – 38.1) | 29.5 (23.2 – 35.9) |
Number of observations in primary analysis after excluding 83 observations with missing covariates. Results corresponding to fellow (n= 9) or other (n=3) clinician type in frontline role are omitted from the table due to low sample size.
Mixed-effects model using random intercepts to account for hospital- and patient-level clustering, without adjustment for covariates
Mixed-effects model using random intercepts to account for hospital- and patient-level clustering, adjusted for patient-level covariates (age combined with preterm birth, time since weaning from supplemental oxygen or flow, documented history of apnea or cyanosis during the present illness, prior intensive care unit stay during present hospitalization, presence of an enteral feeding tube, history of neurologic impairment, and nighttime observation), study phase, intervention arm, and presence of pre-existing clinical decision support (CDS). Results corresponding to alternative models adjusting for hospital type, and both hospital type and clinical decision support, are presented in Appendix Table 2.
DISCUSSION
In this 36-hospital study of over 4,000 cSpO2 observations, cSpO2 monitoring overuse was more likely among patients with frontline care delivered by attendings compared to residents. Contrary to our hypothesis, this finding – robust to confounding by other factors – suggests that trainee involvement in frontline patient care may attenuate cSpO2 overuse. Based on prior work characterizing alarm rates associated with guideline-discordant versus guideline-concordant cSpO2 monitoring (6.7 vs. 0.5 alarms/hour),7 the observed 5.2-percentage point reduction in cSpO2 overuse among resident-led teams translates to 7.7 fewer alarms per patient-day, with the potential to mitigate costs and downstream cascade effects of overuse such as reducing alarm fatigue and length of stay.8
Prior qualitative studies of clinicians have suggested that trainees may be less comfortable deimplementing cSpO2 due to limited clinical experience.9,10 Our data challenge this assumption and suggest alternative explanations. First, resident physicians’ relative inexperience and exposure to diverse care settings during their training may facilitate deimplementation. Because training environments emphasize inquiry and frequent feedback, residents often have less entrenched practices and greater openness to change,11 potentially increasing receptiveness to evidence-based guidelines and new workflows. Second, relative to attendings, residents may have less hierarchical interactions with nurses, which could facilitate nurse-driven discontinuation of cSpO2 monitoring.10 This dynamic may help explain why resident-led teams had lower cSpO2 overuse compared with attending-led teams but not APP-led teams, as APPs may similarly have less hierarchical interactions with nurses. APPs have also demonstrated more favorable attitudes toward reducing practice variation and standardizing care,12 suggesting a professional culture consistent with the principles emphasized in contemporary residency training.13 Third, prior work specific to cSpO2 use in bronchiolitis10 and in other clinical contexts14 has shown that resident engagement in evidence-based practices is highly motivated by the behavior of more senior physicians. This may make residents more responsive to the education and audit-and-feedback strategies embedded in the trial. At the same time, residents’ position within the clinical hierarchy15 relative to the attendings serving as trial site leads may make them less likely to act upon any psychological reactance (resistance to threatened autonomy)16 experienced in response to deimplementation messaging. Future work should formally evaluate these mechanisms and tailor implementation strategies to clinician team structures.
More broadly, this study contributes to a growing literature on the role of distinct inpatient practice models in ensuring quality and mitigating overuse. Multiple systematic reviews and observational studies have demonstrated that teaching services staffed by residents produce equivalent, and at times higher-quality, inpatient care than attending-only teams.17–20 To our knowledge, however, no prior study has specifically examined the association between clinician team structure and care quality in the context of a quality improvement or implementation study. Despite concerns that residents are difficult to engage in quality improvement or have negative attitudes toward such work,21 our findings support the potential value of residents as the foundation of a learning health system.22
Our study has several limitations. First, information on clinician team structure was not recorded during Phase 1 of the trial, limiting our ability to assess baseline differences in the absence of active deimplementation efforts or evaluate longitudinal sustainment of deimplementation.5 Second, clinician team structure was uniform at some hospitals while mixed at others (Appendix Figure). Although we accounted for hospital-level clustering and adjusted for hospital-level variables, unmeasured hospital- or unit-level factors may still influence the association between team structure and cSpO2 overuse. Third, we did not evaluate characteristics of the attending physicians. It is possible that attendings supervising resident teams are different from attendings working in frontline roles with respect to their clinical experience, comfort with reducing cSpO2 monitoring, and length of employment at the hospital, which may influence their alignment with existing institutional practices and culture. Finally, similar to other projects in the EMO portfolio, data collection relied on convenience sampling of cSpO2 use, which may introduce selection bias.
CONCLUSIONS
Inpatient care delivered by teams with resident physicians in a frontline role demonstrated lower cSpO2 overuse than frontline attending teams, a finding that persisted after adjustment for other factors. Future research should investigate the mechanisms by which residents attenuate cSpO2 overuse through mixed-methods inquiry, evaluate whether resident-led teams achieve greater sustainment of overuse reduction, and determine how to tailor implementation strategies to clinician roles. Finally, examining whether these findings extend to other forms of medical overuse will inform broader deimplementation efforts.
Supplementary Material
Funding:
Research reported in this publication was supported in part by a Cooperative Agreement from the National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health (NIH) under award number U01HL159880 (Bonafide and Beidas, MPI). Consistent with the terms of an NIH U01 Cooperative Agreement, NIH scientists participated in study meetings and provided ongoing feedback on the conduct, findings, and interpretation of the trial. The funding organization had no role in the preparation of the manuscript or decision to submit the manuscript for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Additional support for this substudy was provided by Children’s Hospital of Philadelphia and Cleveland Clinic Children’s Hospital.
This manuscript is the result of funding in whole or in part by the National Institutes of Health (NIH). It is subject to the NIH Public Access Policy. Through acceptance of this federal funding, NIH has been given a right to make this manuscript publicly available in PubMed Central upon the Official Date of Publication, as defined by NIH.
Conflicts of Interest:
Dr. Beidas receives funding from the National Institutes of Health, Northwestern Medicine, National Psoriasis Foundation, American Foundation for Suicide Prevention, West Health Institute, Sanofi, and Pfizer for her research. Dr. Beidas is principal at Implementation Science & Practice, LLC and serves as co-Editor-In-Chief for Implementation Science. She is currently an appointed member of the National Advisory Mental Health Council and serves on the scientific advisory board for AIM Youth Mental Health Foundation, Klingenstein Third Generation Foundation, and the American Foundation for Suicide Prevention. She was previously an appointed member of the NASEM study, “Blueprint for a national prevention infrastructure for behavioral health disorders,” served on the scientific and advisory board for Optum Behavioral Health, received consulting fees from United Behavioral Health and OptumLabs, and received royalties from Oxford University Press. All activities are outside of the submitted work. Dr. Enrique Schisterman has received consulting fees from Merck for work on causal inference methods. He is Editor-in-Chief of the American Journal of Epidemiology. None of these activities are directly related to the work presented here.
Footnotes
Clinical Trial Registration: This substudy analysis includes data from ClinicalTrials.gov NCT05132322.
REFERENCES
- 1.Schondelmeyer AC, Dewan ML, Brady PW, et al. Cardiorespiratory and Pulse Oximetry Monitoring in Hospitalized Children: A Delphi Process. Pediatrics. Aug 2020;146(2)doi: 10.1542/peds.2019-3336 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mahant S, Wahi G, Bayliss A, et al. Intermittent vs Continuous Pulse Oximetry in Hospitalized Infants With Stabilized Bronchiolitis: A Randomized Clinical Trial. JAMA Pediatr. May 1 2021;175(5):466–474. doi: 10.1001/jamapediatrics.2020.6141 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bonafide CP, Xiao R, Brady PW, et al. Prevalence of Continuous Pulse Oximetry Monitoring in Hospitalized Children With Bronchiolitis Not Requiring Supplemental Oxygen. JAMA. Apr 21 2020;323(15):1467–1477. doi: 10.1001/jama.2020.2998 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Stoeck PA, Chieco DF, Pingree EW, Landrigan CP, Bonafide CP, Pediatric Research in Inpatient Settings N. Association Between Bronchiolitis Patient Volume and Continuous Pulse Oximetry Monitoring in 25 Hospitals. J Hosp Med. Nov 2020;15(11):669–672. doi: 10.12788/jhm.3520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bonafide CP, Xiao R, Schondelmeyer AC, et al. Sustainable deimplementation of continuous pulse oximetry monitoring in children hospitalized with bronchiolitis: study protocol for the Eliminating Monitor Overuse (EMO) type III effectiveness-deimplementation cluster-randomized trial. Implement Sci. Oct 21 2022;17(1):72. doi: 10.1186/s13012-022-01246-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Xiao R, Bonafide CP, Williams NJ, et al. Eliminating Monitor Overuse (EMO) type III effectiveness-deimplementation cluster-randomized trial: Statistical analysis plan. Contemp Clin Trials Commun. Dec 2023;36:101219. doi: 10.1016/j.conctc.2023.101219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rasooly IR, Makeneni S, Khan AN, Luo B, Muthu N, Bonafide CP. The Alarm Burden of Excess Continuous Pulse Oximetry Monitoring Among Patients With Bronchiolitis. J Hosp Med. Dec 2021;16(12):727–729. doi: 10.12788/jhm.3731 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ganguli I, Lupo C, Mainor AJ, et al. Prevalence and Cost of Care Cascades After Low-Value Preoperative Electrocardiogram for Cataract Surgery in Fee-for-Service Medicare Beneficiaries. JAMA Intern Med. Sep 1 2019;179(9):1211–1219. doi: 10.1001/jamainternmed.2019.1739 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wolk CB, Schondelmeyer AC, Barg FK, et al. Barriers and Facilitators to Guideline-Adherent Pulse Oximetry Use in Bronchiolitis. J Hosp Med. Jan 2021;16(1):23–30. doi: 10.12788/jhm.3535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yama B, Wahi G, Zhou K, et al. De-implementing low-value continuous pulse oximetry practice in infants hospitalized with bronchiolitis: A multicentre qualitative study. J Hosp Med. Dec 2023;18(12):1092–1101. doi: 10.1002/jhm.13236 [DOI] [PubMed] [Google Scholar]
- 11.Montini T, Graham ID. “Entrenched practices and other biases”: unpacking the historical, economic, professional, and social resistance to de-implementation. Implementation Science. 2015/02/13 2015;10(1):24. doi: 10.1186/s13012-015-0211-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cook DA, Pencille LJ, Dupras DM, Linderbaum JA, Pankratz VS, Wilkinson JM. Practice variation and practice guidelines: Attitudes of generalist and specialist physicians, nurse practitioners, and physician assistants. PLoS One. 2018;13(1):e0191943. doi: 10.1371/journal.pone.0191943 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cooper D, Holmboe ES. Competency-Based Medical Education at the Front Lines of Patient Care. N Engl J Med. Jul 24 2025;393(4):376–388. doi: 10.1056/NEJMra2411880 [DOI] [PubMed] [Google Scholar]
- 14.Hong B, O'Sullivan ED, Henein C, Jones CM. Motivators and barriers to engagement with evidence-based practice among medical and dental trainees from the UK and Republic of Ireland: a national survey. BMJ Open. Oct 17 2019;9(10):e031809. doi: 10.1136/bmjopen-2019-031809 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Voogt JJ, Kars MC, van Rensen ELJ, Schneider MME, Noordegraaf M, van der Schaaf MF. Why Medical Residents Do (and Don't) Speak Up About Organizational Barriers and Opportunities to Improve the Quality of Care. Acad Med. Apr 2020;95(4):574–581. doi: 10.1097/ACM.0000000000003014 [DOI] [PubMed] [Google Scholar]
- 16.Dillard JP, Shen L. On the Nature of Reactance and its Role in Persuasive Health Communication. Communication Monographs. 2005/06/01 2005;72(2):144–168. doi: 10.1080/03637750500111815 [DOI] [Google Scholar]
- 17.Au AG, Padwal RS, Majumdar SR, McAlister FA. Patient outcomes in teaching versus nonteaching general internal medicine services: a systematic review and meta-analysis. Acad Med. Mar 2014;89(3):517–23. doi: 10.1097/ACM.0000000000000154 [DOI] [PubMed] [Google Scholar]
- 18.Perez JA Jr., Awar M, Nezamabadi A, et al. Comparison of Direct Patient Care Costs and Quality Outcomes of the Teaching and Nonteaching Hospitalist Services at a Large Academic Medical Center. Acad Med. Mar 2018;93(3):491–497. doi: 10.1097/ACM.0000000000002026 [DOI] [PubMed] [Google Scholar]
- 19.Johnson SA, Ciarkowski CE, Lappe KL, Kendrick DR, Smith A, Reddy SP. Comparison of Resident, Advanced Practice Clinician, and Hospitalist Teams in an Academic Medical Center: Association With Clinical Outcomes and Resource Utilization. J Hosp Med. Dec 2020;15(12):709–715. doi: 10.12788/jhm.3475 [DOI] [PubMed] [Google Scholar]
- 20.Darby JB, Singhal G, Halvorson EE, et al. Comparison of a Teaching and Nonteaching Service at a Children's Hospital in the Community. Acad Pediatr. Apr 2022;22(3):440–446. doi: 10.1016/j.acap.2021.07.004 [DOI] [PubMed] [Google Scholar]
- 21.Butler JM, Anderson KA, Supiano MA, Weir CR. “It Feels Like a Lot of Extra Work”: Resident Attitudes About Quality Improvement and Implications for an Effective Learning Health Care System. Acad Med. Jul 2017;92(7):984–990. doi: 10.1097/ACM.0000000000001474 [DOI] [PubMed] [Google Scholar]
- 22.Lam PW, Wong BM. Harnessing the Power of Residents as Change Agents in Quality Improvement. Acad Med. Jan 1 2021;96(1):21–23. doi: 10.1097/ACM.0000000000003727 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.