Nathan Mesfin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Nathan Mesfin has a total of 74 co-authors. Here are the top co-authors by number of shared publications:
Benjamin Langworthy
3 shared publications- Nicholas E. Ingraham
University of Minnesota Medical School; University of Minnesota Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine Minneapolis, MN
3 shared publications - R. Adams Dudley
Pulmonary, Allergy, Critical Care, and Sleep Medicine, University of Minnesota, Minneapolis, MN, United States
3 shared publications Renda Soylemez Wiener
Affiliations Center for Healthcare Organization and Implementation Research (CHOIR), VA Boston Healthcare System, Boston, Massachusetts, United States of America, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, United States of America, Department of Medicine, VA Boston Healthcare System, Boston, Massachusetts, United States of America
3 shared publicationsAmy M. Linsky
Affiliations Center for Healthcare Organization and Implementation Research (CHOIR), VA Boston Healthcare System, Boston, Massachusetts, United States of America, Department of Medicine, Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts, United States of America, Department of Medicine, VA Boston Healthcare System, Boston, Massachusetts, United States of America
2 shared publicationsChris S. Xu
Affiliation Center for Healthcare Organization and Implementation Research (CHOIR), VA Boston Healthcare System, Boston, Massachusetts, United States of America
2 shared publications
Latest Publications
- Open Access
- 13 October 2026
- Published by Elsevier in CHEST Critical Care
Provider Variation in Time to Do Not Resuscitate Decisions and Its Association With Mortality
- 13 October 2026
- Published by Elsevier in CHEST Critical Care
Open AccessProvider Variation in Time to Do Not Resuscitate Decisions and Its Association With Mortality
- Open Access
- 9 September 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e97485
Natural Language Processing Identification of Nonprescribed Fentanyl Use in Electronic Health Records: Algorithm Development and Validation Study
Background: Overdose and suicide due to nonprescribed fentanyl use have increased significantly, yet health care systems lack reliable methods to identify patients who use nonprescribed fentanyl. International Classification of Diseases codes are inconsistent and do not specify nonprescribed fentanyl use. Objective: This study aimed to develop natural language processing approaches to identifying nonprescribed fentanyl use in electronic health record (EHR) documentation. Methods: This retrospective study included Veterans Health Administration patients seen between April 5, 2023, and December 23, 2024. A term list was developed to identify fentanyl-related mentions in clinical text, and 250-character snippets surrounding identified mentions were extracted. Veterans (n=3878) were randomly sampled from 5 predefined groups based on the presence of 1 of 4 terms (“fent,” “blues,” “M30s,” and “tranq”) in their EHR documentation. Physician annotators classified snippets into “nonprescribed fentanyl use,” “prescribed fentanyl use,” or “other,” with interannotator agreement evaluated using the mean pairwise Cohen κ. Cross-validation folds were constructed at the patient level between training and test sets. Penalized logistic regression, Bio-ClinicalBERT, Llama 3-8B, and Mistral-7B were trained on labeled data and compared. Model performance was evaluated using precision, recall, and F 1 -scores for each class, with a focus on the nonprescribed fentanyl use class as the primary label of clinical interest using bootstrapped 95% CIs. A fairness analysis and Shapley additive explanations analysis were performed using Bio-ClinicalBERT. External validation was performed using Bio-ClinicalBERT on an independent sample of 200 snippets, each representing a unique patient from January 2025 to June 2026, with precision reported as the primary validation metric. Results: Of 7389 snippets, 9.6% (n=709) were classified as “nonprescribed fentanyl use,” 40.3% (n=2981) were classified as “prescribed fentanyl use,” and 50% (n=3699) were classified as “other.” Interannotator agreement was high (κ=0.822). Llama 3-8B achieved the highest F 1 -score for nonprescribed fentanyl use (0.87, 95% CI 0.83-0.92), followed by Mistral-7B (0.80, 95% CI 0.75-0.84), Bio-ClinicalBERT (0.80, 95% CI 0.74-0.85), and penalized logistic regression (0.74, 95% CI 0.73-0.75). Performance was consistent across demographic subgroups, with lower performance for the nonprescribed fentanyl use class observed in female and Hispanic subgroups. Shapley additive explanations analysis revealed clinically meaningful discriminating terms for each class, although subword tokens required contextual interpretation. External validation of Bio-ClinicalBERT demonstrated a precision of 0.79 for nonprescribed fentanyl use. Conclusions: Natural language processing can identify nonprescribed fentanyl use in EHR documentation, although model performance for this class was lower than overall model performance, reflecting the clinical complexity of identifying nonprescribed use and the variable ways in which clinicians document this problem. This approach may support risk prediction and targeting of interventions to patients exposed to nonprescribed fentanyl.
- 9 September 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e97485
Open AccessNatural Language Processing Identification of Nonprescribed Fentanyl Use in Electronic Health Records: Algorithm Development and Validation Study
Background: Overdose and suicide due to nonprescribed fentanyl use have increased significantly, yet health care systems lack reliable methods to identify patients who use nonprescribed fentanyl. International Classification of Diseases codes are inconsistent and do not specify nonprescribed fentanyl use. Objective: This study aimed to develop natural language processing approaches to identifying nonprescribed fentanyl use in electronic health record (EHR) documentation. Methods: This retrospective study included Veterans Health Administration patients seen between April 5, 2023, and December 23, 2024. A term list was developed to identify fentanyl-related mentions in clinical text, and 250-character snippets surrounding identified mentions were extracted. Veterans (n=3878) were randomly sampled from 5 predefined groups based on the presence of 1 of 4 terms (“fent,” “blues,” “M30s,” and “tranq”) in their EHR documentation. Physician annotators classified snippets into “nonprescribed fentanyl use,” “prescribed fentanyl use,” or “other,” with interannotator agreement evaluated using the mean pairwise Cohen κ. Cross-validation folds were constructed at the patient level between training and test sets. Penalized logistic regression, Bio-ClinicalBERT, Llama 3-8B, and Mistral-7B were trained on labeled data and compared. Model performance was evaluated using precision, recall, and F 1 -scores for each class, with a focus on the nonprescribed fentanyl use class as the primary label of clinical interest using bootstrapped 95% CIs. A fairness analysis and Shapley additive explanations analysis were performed using Bio-ClinicalBERT. External validation was performed using Bio-ClinicalBERT on an independent sample of 200 snippets, each representing a unique patient from January 2025 to June 2026, with precision reported as the primary validation metric. Results: Of 7389 snippets, 9.6% (n=709) were classified as “nonprescribed fentanyl use,” 40.3% (n=2981) were classified as “prescribed fentanyl use,” and 50% (n=3699) were classified as “other.” Interannotator agreement was high (κ=0.822). Llama 3-8B achieved the highest F 1 -score for nonprescribed fentanyl use (0.87, 95% CI 0.83-0.92), followed by Mistral-7B (0.80, 95% CI 0.75-0.84), Bio-ClinicalBERT (0.80, 95% CI 0.74-0.85), and penalized logistic regression (0.74, 95% CI 0.73-0.75). Performance was consistent across demographic subgroups, with lower performance for the nonprescribed fentanyl use class observed in female and Hispanic subgroups. Shapley additive explanations analysis revealed clinically meaningful discriminating terms for each class, although subword tokens required contextual interpretation. External validation of Bio-ClinicalBERT demonstrated a precision of 0.79 for nonprescribed fentanyl use. Conclusions: Natural language processing can identify nonprescribed fentanyl use in EHR documentation, although model performance for this class was lower than overall model performance, reflecting the clinical complexity of identifying nonprescribed use and the variable ways in which clinicians document this problem. This approach may support risk prediction and targeting of interventions to patients exposed to nonprescribed fentanyl.
- Open Access
- 7 April 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Natural Language Processing Identification of Non-prescribed Fentanyl Use in Electronic Health Records (Preprint)
BACKGROUND Overdose and suicide deaths due to non-prescribed fentanyl have increased significantly. Fortunately, there are treatments available that could reduce the risk of death. However, for healthcare systems to implement programs addressing the needs of patients who use non-prescribed fentanyl, they must first be able to identify them. International Classification of Diseases (ICD) codes are not a good option for identifying patients who use non-prescribed fentanyl for two reasons. First, opioid use disorder (OUD) diagnoses are not always coded. Second, when they are coded, ICD codes for OUD do not specify whether fentanyl is being used rather than other, less lethal opioids. OBJECTIVE To develop natural language processing (NLP) approaches to identifying current non-prescribed fentanyl use in electronic health record (EHR) documentation. METHODS This retrospective cross-sectional study included Veterans Health Administration (VHA) patients seen in the VHA between 4/5/23 and 12/23/24. A term list was developed to identify fentanyl-related mentions in clinical text, and 250-character snippets surrounding identified mentions were extracted. Veterans (n=3,878) were randomly sampled from five predefined groups based on the presence of one of four terms (fentanyl, fent, blues, tranq) in their EHR documentation. Physician annotators classified snippets into “non-prescribed fentanyl use,” “prescribed fentanyl use,” or “other.” Labeled data was used to train, validate, and compare the ability of penalized logistic regression, Bio-Clinical BERT, LLaMA 3 8B, and Mistral 7B to classify snippets. Model performance was evaluated using weighted average precision, recall, and F1 scores. RESULTS Results: Among 7,389 snippets, 9.6% were “non-prescribed fentanyl use,” 40.3% were “prescribed fentanyl use,” and 50% were “other.” LLaMA 3 8B achieved the highest weighted average F1 score (0.954) outperformed penalized logistic regression with a small statistically significant difference, while differences between Mistral 7B and LLaMA 3 8B were not statistically significant. Top discriminative terms included “consumption,” “purchased,” and “streets” for non prescribed fentanyl use; “dressing,” “patch(es),” and “1500” for prescribed; and “laced,” “testing,” and “fears” for other fentanyl use. CONCLUSIONS NLP can accurately identify non-prescribed fentanyl use in EHR documentation. This approach may support risk prediction and targeting of interventions to patients exposed to non-prescribed fentanyl.
- 7 April 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessNatural Language Processing Identification of Non-prescribed Fentanyl Use in Electronic Health Records (Preprint)
BACKGROUND Overdose and suicide deaths due to non-prescribed fentanyl have increased significantly. Fortunately, there are treatments available that could reduce the risk of death. However, for healthcare systems to implement programs addressing the needs of patients who use non-prescribed fentanyl, they must first be able to identify them. International Classification of Diseases (ICD) codes are not a good option for identifying patients who use non-prescribed fentanyl for two reasons. First, opioid use disorder (OUD) diagnoses are not always coded. Second, when they are coded, ICD codes for OUD do not specify whether fentanyl is being used rather than other, less lethal opioids. OBJECTIVE To develop natural language processing (NLP) approaches to identifying current non-prescribed fentanyl use in electronic health record (EHR) documentation. METHODS This retrospective cross-sectional study included Veterans Health Administration (VHA) patients seen in the VHA between 4/5/23 and 12/23/24. A term list was developed to identify fentanyl-related mentions in clinical text, and 250-character snippets surrounding identified mentions were extracted. Veterans (n=3,878) were randomly sampled from five predefined groups based on the presence of one of four terms (fentanyl, fent, blues, tranq) in their EHR documentation. Physician annotators classified snippets into “non-prescribed fentanyl use,” “prescribed fentanyl use,” or “other.” Labeled data was used to train, validate, and compare the ability of penalized logistic regression, Bio-Clinical BERT, LLaMA 3 8B, and Mistral 7B to classify snippets. Model performance was evaluated using weighted average precision, recall, and F1 scores. RESULTS Results: Among 7,389 snippets, 9.6% were “non-prescribed fentanyl use,” 40.3% were “prescribed fentanyl use,” and 50% were “other.” LLaMA 3 8B achieved the highest weighted average F1 score (0.954) outperformed penalized logistic regression with a small statistically significant difference, while differences between Mistral 7B and LLaMA 3 8B were not statistically significant. Top discriminative terms included “consumption,” “purchased,” and “streets” for non prescribed fentanyl use; “dressing,” “patch(es),” and “1500” for prescribed; and “laced,” “testing,” and “fears” for other fentanyl use. CONCLUSIONS NLP can accurately identify non-prescribed fentanyl use in EHR documentation. This approach may support risk prediction and targeting of interventions to patients exposed to non-prescribed fentanyl.
- 23 January 2026
- Published by Oxford University Press (OUP) in Annals of the American Thoracic Society
- Vol. 23 (4) , 475-482
Federation, not centralization: a new paradigm for electronic health record–based critical care research
- 23 January 2026
- Published by Oxford University Press (OUP) in Annals of the American Thoracic Society
- Vol. 23 (4) , 475-482
Federation, not centralization: a new paradigm for electronic health record–based critical care research
- Open Access
- 1 June 2025
- Published by SAGE Publications in Palliative Medicine Reports
- Vol. 6 (1) , 468-473
Hospital and Temporal Variations in Limitations of Care Among Hospitalized Patients with COVID-19: A VIRUS Registry Retrospective Cohort Study
Introduction: Life-support limits (i.e., code status), the first step in the continuum of palliation and end-of-life (EoL) care, stem from decisions made by patients and their providers to align care near the EoL with patients’ wishes. The pandemic created health care strain that may have influenced life-support limitation practices among patients with serious COVID-19. In this study, we examine variations in life-support limitations (do-not-resuscitate, DNR; do-not-intubate, DNI) depending on the hospital and period of hospitalization. Methods: We included all adults admitted to a hospital in the United States with COVID-19 from January 2020 to December 2021 using the VIRUS registry. Our outcome was any life-support limitation on admission (DNR, DNI, or both) using regression modeling. Main exposures were hospital and period of hospitalization (early vs. late pandemic). Covariates included age, sex, race/ethnicity, comorbidities, and hospitalization diagnoses. Results: There were 42,383 patients from 75 hospitals in the US. The median age was 63 (interquartile range: 50–75), 46.5% were female, and 37.5% were non-Hispanic White. The life-support limitation rate was 7.9%, and the palliative care consultation rate was 2.3%. The odds of life-support limitation were 1.86 (95% CI: 1.6–2.1) depending on the hospital, but the period of hospitalization was not associated with odds of life-support limitation. Older age was associated with the greatest increased odds (4.1 per standard deviation ∼ 17.9 years, 95% CI: 3.9–4.4), followed by comorbidities including paralysis (2.4, 95% CI: 1.8–3.0) and dementia (2.3, 95% CI: 2.0–2.6). Conclusions: After adjusting for patient-level factors, we report significant inter-hospital but no temporal variation in life-support limitation upon admission with COVID-19. Future studies should investigate specific practices that led to health care resilience to temporal practice variations in life-support limitation during the pandemic.
- 1 June 2025
- Published by SAGE Publications in Palliative Medicine Reports
- Vol. 6 (1) , 468-473
Open AccessHospital and Temporal Variations in Limitations of Care Among Hospitalized Patients with COVID-19: A VIRUS Registry Retrospective Cohort Study
Introduction: Life-support limits (i.e., code status), the first step in the continuum of palliation and end-of-life (EoL) care, stem from decisions made by patients and their providers to align care near the EoL with patients’ wishes. The pandemic created health care strain that may have influenced life-support limitation practices among patients with serious COVID-19. In this study, we examine variations in life-support limitations (do-not-resuscitate, DNR; do-not-intubate, DNI) depending on the hospital and period of hospitalization. Methods: We included all adults admitted to a hospital in the United States with COVID-19 from January 2020 to December 2021 using the VIRUS registry. Our outcome was any life-support limitation on admission (DNR, DNI, or both) using regression modeling. Main exposures were hospital and period of hospitalization (early vs. late pandemic). Covariates included age, sex, race/ethnicity, comorbidities, and hospitalization diagnoses. Results: There were 42,383 patients from 75 hospitals in the US. The median age was 63 (interquartile range: 50–75), 46.5% were female, and 37.5% were non-Hispanic White. The life-support limitation rate was 7.9%, and the palliative care consultation rate was 2.3%. The odds of life-support limitation were 1.86 (95% CI: 1.6–2.1) depending on the hospital, but the period of hospitalization was not associated with odds of life-support limitation. Older age was associated with the greatest increased odds (4.1 per standard deviation ∼ 17.9 years, 95% CI: 3.9–4.4), followed by comorbidities including paralysis (2.4, 95% CI: 1.8–3.0) and dementia (2.3, 95% CI: 2.0–2.6). Conclusions: After adjusting for patient-level factors, we report significant inter-hospital but no temporal variation in life-support limitation upon admission with COVID-19. Future studies should investigate specific practices that led to health care resilience to temporal practice variations in life-support limitation during the pandemic.
- Open Access
- 1 January 2025
- Published by Wiley in Pulmonary Circulation
- Vol. 15 (1) , e70021
Contribution of QRS Prolongation to Transplant‐Free Survival in Sarcoidosis‐Related Pulmonary Hypertension
To study the contribution of QRS prolongation to transplant‐free survival, we conducted an observational study of 68 patients with sarcoidosis‐related pulmonary hypertension. Every 10‐ms increase in QRS interval was associated with an adjusted HR of 1.23 (95% CI: 1.06, 1.44) for the composite outcome of lung transplantation and all‐cause mortality.
- 1 January 2025
- Published by Wiley in Pulmonary Circulation
- Vol. 15 (1) , e70021
Open AccessContribution of QRS Prolongation to Transplant‐Free Survival in Sarcoidosis‐Related Pulmonary Hypertension
To study the contribution of QRS prolongation to transplant‐free survival, we conducted an observational study of 68 patients with sarcoidosis‐related pulmonary hypertension. Every 10‐ms increase in QRS interval was associated with an adjusted HR of 1.23 (95% CI: 1.06, 1.44) for the composite outcome of lung transplantation and all‐cause mortality.
- Open Access
- 15 September 2023
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 18 (9) , e0291542
Addressing clinician moral distress: Implications from a mixed methods evaluation during Covid-19
Clinician moral distress has been documented over the past several decades as occurring within numerous healthcare disciplines, often in relation to clinicians’ involvement in patients’ end-of-life decision-making. The resulting harms impact clinician well-being, patient well-being, and healthcare system functioning. Given Covid-19’s catastrophic death toll and associated demands on end-of-life decision-making processes, the pandemic represents a particularly important context within which to understand clinician moral distress. Thus, we conducted a convergent mixed methods study to examine its prevalence, associations with clinicians’ demographic and professional characteristics, and contributing circumstances among Veterans Health Administration (VA) clinicians. The study, conducted in April 2021, consisted of a cross-sectional on-line survey of VA clinicians at 20 VA Medical Centers with professional jurisdiction to place life-sustaining treatment orders working who were from a number of select specialties. The survey collected quantitative data on respondents’ demographics, clinical practice characteristics, attitudes and behaviors related to goals of care conversations, intensity of moral distress during “peak-Covid,” and qualitative data via an open-ended item asking for respondents to describe contributing circumstances if they had indicated any moral distress. To understand factors associated with heightened moral distress, we analyzed quantitative data using bivariate and multivariable regression analyses and qualitative data using a hybrid deductive/inductive thematic approach. Mixed methods analysis followed, whereby we compared the quantitative and qualitative datasets and integrated findings at the analytic level. Out of 3,396 eligible VA clinicians, 323 responded to the survey (9.5% adjusted response rate). Most respondents (81%) reported at least some moral distress during peak-Covid. In a multivariable logistic regression, female gender (OR 3.35; 95% CI 1.53–7.37) was associated with greater odds of moral distress, and practicing in geriatrics/palliative care (OR 0.40; 95% CI 0.18–0.87) and internal medicine/family medicine/primary care (OR 0.46; 95% CI 0.22–0.98) were associated with reduced odds of moral distress compared to medical subspecialties. From the 191 respondents who completed the open-ended item, five qualitative themes emerged as moral distress contributors: 1) patient visitation restrictions, 2) anticipatory actions, 3) clinical uncertainty related to Covid, 4) resource shortages, and 5) personal risk of contracting Covid. Mixed methods analysis found that quantitative results were consistent with these last two qualitative themes. In sum, clinician moral distress was prevalent early in the pandemic. This moral distress was associated with individual-, system-, and situation-level contributors. These identified contributors represent leverage points for future intervention to mitigate clinician moral distress and its negative outcomes during future healthcare crises and even during everyday clinical care.
- 15 September 2023
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 18 (9) , e0291542
Open AccessAddressing clinician moral distress: Implications from a mixed methods evaluation during Covid-19
Clinician moral distress has been documented over the past several decades as occurring within numerous healthcare disciplines, often in relation to clinicians’ involvement in patients’ end-of-life decision-making. The resulting harms impact clinician well-being, patient well-being, and healthcare system functioning. Given Covid-19’s catastrophic death toll and associated demands on end-of-life decision-making processes, the pandemic represents a particularly important context within which to understand clinician moral distress. Thus, we conducted a convergent mixed methods study to examine its prevalence, associations with clinicians’ demographic and professional characteristics, and contributing circumstances among Veterans Health Administration (VA) clinicians. The study, conducted in April 2021, consisted of a cross-sectional on-line survey of VA clinicians at 20 VA Medical Centers with professional jurisdiction to place life-sustaining treatment orders working who were from a number of select specialties. The survey collected quantitative data on respondents’ demographics, clinical practice characteristics, attitudes and behaviors related to goals of care conversations, intensity of moral distress during “peak-Covid,” and qualitative data via an open-ended item asking for respondents to describe contributing circumstances if they had indicated any moral distress. To understand factors associated with heightened moral distress, we analyzed quantitative data using bivariate and multivariable regression analyses and qualitative data using a hybrid deductive/inductive thematic approach. Mixed methods analysis followed, whereby we compared the quantitative and qualitative datasets and integrated findings at the analytic level. Out of 3,396 eligible VA clinicians, 323 responded to the survey (9.5% adjusted response rate). Most respondents (81%) reported at least some moral distress during peak-Covid. In a multivariable logistic regression, female gender (OR 3.35; 95% CI 1.53–7.37) was associated with greater odds of moral distress, and practicing in geriatrics/palliative care (OR 0.40; 95% CI 0.18–0.87) and internal medicine/family medicine/primary care (OR 0.46; 95% CI 0.22–0.98) were associated with reduced odds of moral distress compared to medical subspecialties. From the 191 respondents who completed the open-ended item, five qualitative themes emerged as moral distress contributors: 1) patient visitation restrictions, 2) anticipatory actions, 3) clinical uncertainty related to Covid, 4) resource shortages, and 5) personal risk of contracting Covid. Mixed methods analysis found that quantitative results were consistent with these last two qualitative themes. In sum, clinician moral distress was prevalent early in the pandemic. This moral distress was associated with individual-, system-, and situation-level contributors. These identified contributors represent leverage points for future intervention to mitigate clinician moral distress and its negative outcomes during future healthcare crises and even during everyday clinical care.
- 1 July 2023
- Published by SAGE Publications in Journal of Palliative Medicine
- Vol. 26 (7) , 951-959
Impact of the COVID-19 Pandemic on Providing Recommendations During Goals-of-Care Conversations: A Multisite Survey
Background: Goals-of-care conversations (GoCCs) are essential for individualized end-of-life care. Shared decision-making (SDM) that elicits patients' goals and values to collaboratively make life sustaining treatment (LST) decisions is best practice. However, it is unknown how the COVID-19 pandemic onset and associated changes to care delivery, stress on providers, and clinical uncertainty affected SDM and recommendation-making during GoCCs. Aim: To assess providers' attitudes and behaviors related to GoCCs during the COVID-19 pandemic and identify factors associated with provision of LST recommendations. Design: Survey of United States Veterans Health Administration (VA) health care providers. Setting/Participants: Health care providers from 20 VA facilities with high COVID-19 caseloads early in the pandemic who had authority to place LST orders and practiced in select specialties (n = 3398). Results: We had 323 respondents (9.5% adjusted response rate). Most were age ≥50 years (51%), female (63%), non-Hispanic white (64%), and had ≥1 GoCC per week during peak-COVID-19 (78%). Compared with pre-COVID-19, providers believed it was less appropriate and felt less comfortable giving an LST recommendation during peak-COVID-19 (p < 0.001). One-third (32%) reported either “never” or “rarely” giving an LST recommendation during GoCCs at peak-COVID-19. In adjusted regression models, being a physician and discussing patients' goals and values were positively associated with giving an LST recommendation (B = 0.380, p = 0.031 and B = 0.400, p < 0.001, respectively) at peak-COVID-19. Conclusion: Providers who discuss patients' preferences and values are more likely to report giving a recommendation; both behaviors are markers of SDM during GoCCs. Our findings suggest potential areas for training in conducting patient-centered GoCCs.
- 1 July 2023
- Published by SAGE Publications in Journal of Palliative Medicine
- Vol. 26 (7) , 951-959
Impact of the COVID-19 Pandemic on Providing Recommendations During Goals-of-Care Conversations: A Multisite Survey
Background: Goals-of-care conversations (GoCCs) are essential for individualized end-of-life care. Shared decision-making (SDM) that elicits patients' goals and values to collaboratively make life sustaining treatment (LST) decisions is best practice. However, it is unknown how the COVID-19 pandemic onset and associated changes to care delivery, stress on providers, and clinical uncertainty affected SDM and recommendation-making during GoCCs. Aim: To assess providers' attitudes and behaviors related to GoCCs during the COVID-19 pandemic and identify factors associated with provision of LST recommendations. Design: Survey of United States Veterans Health Administration (VA) health care providers. Setting/Participants: Health care providers from 20 VA facilities with high COVID-19 caseloads early in the pandemic who had authority to place LST orders and practiced in select specialties (n = 3398). Results: We had 323 respondents (9.5% adjusted response rate). Most were age ≥50 years (51%), female (63%), non-Hispanic white (64%), and had ≥1 GoCC per week during peak-COVID-19 (78%). Compared with pre-COVID-19, providers believed it was less appropriate and felt less comfortable giving an LST recommendation during peak-COVID-19 (p < 0.001). One-third (32%) reported either “never” or “rarely” giving an LST recommendation during GoCCs at peak-COVID-19. In adjusted regression models, being a physician and discussing patients' goals and values were positively associated with giving an LST recommendation (B = 0.380, p = 0.031 and B = 0.400, p < 0.001, respectively) at peak-COVID-19. Conclusion: Providers who discuss patients' preferences and values are more likely to report giving a recommendation; both behaviors are markers of SDM during GoCCs. Our findings suggest potential areas for training in conducting patient-centered GoCCs.
- Open Access
- 4 May 2023
- Published by Elsevier in CHEST Pulmonary
- Vol. 1 (1)
Variation in Incidence of Pneumothorax Complications After Percutaneous Lung Biopsies Among Veterans Health Administration Hospitals
- 4 May 2023
- Published by Elsevier in CHEST Pulmonary
- Vol. 1 (1)
Open AccessVariation in Incidence of Pneumothorax Complications After Percutaneous Lung Biopsies Among Veterans Health Administration Hospitals