Health Policy

Artificial intelligence (AI), first described in 1956, encompasses the field of computer science in which machines are trained to learn from experience. The term was popularized by the 1956 Dartmouth College Summer Research Project on Artificial Intelligence.¹ The field of AI is rapidly growing and has the potential to affect many aspects of our lives. The emerging importance of AI is demonstrated by a February 2019 executive order that launched the American AI Initiative, allocating resources and funding for AI development.² The executive order stresses the potential impact of AI in the health care field, including its potential utility to diagnose disease. Federal agencies were directed to invest in AI research and development to promote rapid breakthroughs in AI technology that may impact multiple areas of society.

Machine learning (ML), a subset of AI, was defined in 1959 by Arthur Samuel and is achieved by employing mathematic models to compute sample data sets.³ Originating from statistical linear models, neural networks were conceived to accomplish these tasks.⁴ These pioneering scientific achievements led to recent developments of deep neural networks. These models are developed to recognize patterns and achieve complex computational tasks within a matter of minutes, often far exceeding human ability.⁵ ML can increase efficiency with decreased computation time, high precision, and recall when compared with that of human decision making.⁶

ML has the potential for numerous applications in the health care field.^7-9 One promising application is in the field of anatomic pathology. ML allows representative images to be used to train a computer to recognize patterns from labeled photographs. Based on a set of images selected to represent a specific tissue or disease process, the computer can be trained to evaluate and recognize new and unique images from patients and render a diagnosis.¹⁰ Prior to modern ML models, users would have to import many thousands of training images to produce algorithms that could recognize patterns with high accuracy. Modern ML algorithms allow for a model known as transfer learning, such that far fewer images are required for training.^11-13

Two novel ML platforms available for public use are offered through Google (Mountain View, CA) and Apple (Cupertino, CA).^14,15 They each offer a user-friendly interface with minimal experience required in computer science. Google AutoML uses ML via cloud services to store and retrieve data with ease. No coding knowledge is required. The Apple Create ML Module provides computer-based ML, requiring only a few lines of code.

The Veterans Health Administration (VHA) is the largest single health care system in the US, and nearly 50 000 cancer cases are diagnosed at the VHA annually.¹⁶ Cancers of the lung and colon are among the most common sources of invasive cancer and are the 2 most common causes of cancer deaths in America.¹⁶ We have previously reported using Apple ML in detecting non-small cell lung cancers (NSCLCs), including adenocarcinomas and squamous cell carcinomas (SCCs); and colon cancers with accuracy.^17,18 In the present study, we expand on these findings by comparing Apple and Google ML platforms in a variety of common pathologic scenarios in veteran patients. Using limited training data, both programs are compared for precision and recall in differentiating conditions involving lung and colon pathology.

In the first 4 experiments, we evaluated the ability of the platforms to differentiate normal lung tissue from cancerous lung tissue, to distinguish lung adenocarcinoma from SCC, and to differentiate colon adenocarcinoma from normal colon tissue. Next, cases of colon adenocarcinoma were assessed to determine whether the presence or absence of the KRAS proto-oncogene could be determined histologically using the AI platforms. KRAS is found in a variety of cancers, including about 40% of colon adenocarcinomas.¹⁹ For colon cancers, the presence or absence of the mutation in KRAS has important implications for patients as it determines whether the tumor will respond to specific chemotherapy agents.²⁰ The presence of the KRAS gene is currently determined by complex molecular testing of tumor tissue.²¹ However, we assessed the potential of ML to determine whether the mutation is present by computerized morphologic analysis alone. Our last experiment examined the ability of the Apple and Google platforms to differentiate between adenocarcinomas of lung origin vs colon origin. This has potential utility in determining the site of origin of metastatic carcinoma.²²

Methods

Fifty cases of lung SCC, 50 cases of lung adenocarcinoma, and 50 cases of colon adenocarcinoma were randomly retrieved from our molecular database. Twenty-five colon adenocarcinoma cases were positive for mutation in KRAS, while 25 cases were negative for mutation in KRAS. Seven hundred fifty total images of lung tissue (250 benign lung tissue, 250 lung adenocarcinomas, and 250 lung SCCs) and 500 total images of colon tissue (250 benign colon tissue and 250 colon adenocarcinoma) were obtained using a Leica Microscope MC190 HD Camera (Wetzlar, Germany) connected to an Olympus BX41 microscope (Center Valley, PA) and the Leica Acquire 9072 software for Apple computers. All the images were captured at a resolution of 1024 x 768 pixels using a 60x dry objective. Lung tissue images were captured and saved on a 2012 Apple MacBook Pro computer, and colon images were captured and saved on a 2011 Apple iMac computer. Both computers were running macOS v10.13.

Creating Image Classifier Models Using Apple Create ML

Apple Create ML is a suite of products that use various tools to create and train custom ML models on Apple computers.¹⁵ The suite contains many features, including image classification to train a ML model to classify images, natural language processing to classify natural language text, and tabular data to train models that deal with labeling information or estimating new quantities. We used Create ML Image Classification to create image classifier models for our project (Appendix A).

Creating ML Modules Using Google Cloud AutoML Vision Beta

Google Cloud AutoML is a suite of machine learning products, including AutoML Vision, AutoML Natural Language and AutoML Translation.¹⁴ All Cloud AutoML machine learning products were in beta version at the time of experimentation. We used Cloud AutoML Vision beta to create ML modules for our project. Unlike Apple Create ML, which is run on a local Apple computer, the Google Cloud AutoML is run online using a Google Cloud account. There are no minimum specifications requirements for the local computer since it is using the cloud-based architecture (Appendix B).

Experiment 1

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect and subclassify NSCLC based on the histopathologic images. We created 3 classes of images (250 images each): benign lung tissue, lung adenocarcinoma, and lung SCC.

Experiment 2

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between normal lung tissue and NSCLC histopathologic images with 50/50 mixture of lung adenocarcinoma and lung SCC. We created 2 classes of images (250 images each): benign lung tissue and lung NSCLC.

Experiment 3

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and lung SCC histopathologic images. We created 2 classes of images (250 images each): adenocarcinoma and SCC.

Experiment 4

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect colon cancer histopathologic images regardless of mutation in KRAS status. We created 2 classes of images (250 images each): benign colon tissue and colon adenocarcinoma.

Experiment 5

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between colon adenocarcinoma with mutations in KRAS and colon adenocarcinoma without the mutation in KRAS histopathologic images. We created 2 classes of images (125 images each): colon adenocarcinoma cases with mutation in KRAS and colon adenocarcinoma cases without the mutation in KRAS.

Experiment 6

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and colon adenocarcinoma histopathologic images. We created 2 classes of images (250 images each): colon adenocarcinoma lung adenocarcinoma.

Results

Twelve machine learning models were created in 6 experiments using the Apple Create ML and the Google AutoML (Table). To investigate recall and precision differences between the Apple and the Google ML algorithms, we performed 2-tailed distribution, paired t tests. No statistically significant differences were found (P = .52 for recall and .60 for precision).

Overall, each model performed well in distinguishing between normal and neoplastic tissue for both lung and colon cancers. In subclassifying NSCLC into adenocarcinoma and SCC, the models were shown to have high levels of precision and recall. The models also were successful in distinguishing between lung and colonic origin of adenocarcinoma (Figures 1-4). However, both systems had trouble discerning colon adenocarcinoma with mutations in KRAS from adenocarcinoma without mutations in KRAS.

Discussion

Image classifier models using ML algorithms hold a promising future to revolutionize the health care field. ML products, such as those modules offered by Apple and Google, are easy to use and have a simple graphic user interface to allow individuals to train models to perform humanlike tasks in real time. In our experiments, we compared multiple algorithms to determine their ability to differentiate and subclassify histopathologic images with high precision and recall using common scenarios in treating veteran patients.

Analysis of the results revealed high precision and recall values illustrating the models’ ability to differentiate and detect benign lung tissue from lung SCC and lung adenocarcinoma in ML model 1, benign lung from NSCLC carcinoma in ML model 2, and benign colon from colonic adenocarcinoma in ML model 4. In ML model 3 and 6, both ML algorithms performed at a high level to differentiate lung SCC from lung adenocarcinoma and lung adenocarcinoma from colonic adenocarcinoma, respectively. Of note, ML model 5 had the lowest precision and recall values across both algorithms demonstrating the models’ limited utility in predicting molecular profiles, such as mutations in KRAS as tested here. This is not surprising as pathologists currently require complex molecular tests to detect mutations in KRAS reliably in colon cancer.

Both modules require minimal programming experience and are easy to use. In our comparison, we demonstrated critical distinguishing characteristics that differentiate the 2 products.

Apple Create ML image classifier is available for use on local Mac computers that use Xcode version 10 and macOS 10.14 or later, with just 3 lines of code required to perform computations. Although this product is limited to Apple computers, it is free to use, and images are stored on the computer hard drive. Of unique significance on the Apple system platform, images can be augmented to alter their appearance to enhance model training. For example, imported images can be cropped, rotated, blurred, and flipped, in order to optimize the model’s training abilities to recognize test images and perform pattern recognition. This feature is not as readily available on the Google platform. Apple Create ML Image classifier’s default training set consists of 75% of total imported images with 5% of the total images being randomly used as a validation set. The remaining 20% of images comprise the testing set. The module’s computational analysis to train the model is achieved in about 2 minutes on average. The score threshold is set at 50% and cannot be manipulated for each image class as in Google AutoML Vision.

Google AutoML Vision is open and can be accessed from many devices. It stores images on remote Google servers but requires computing fees after a $300 credit for 12 months. On AutoML Vision, random 80% of the total images are used in the training set, 10% are used in the validation set, and 10% are used in the testing set. It is important to highlight the different percentages used in the default settings on the respective modules. The time to train the Google AutoML Vision with default computational power is longer on average than Apple Create ML, with about 8 minutes required to train the machine learning module. However, it is possible to choose more computational power for an additional fee and decrease module training time. The user will receive e-mail alerts when the computer time begins and is completed. The computation time is calculated by subtracting the time of the initial e-mail from the final e-mail.

Based on our calculations, we determined there was no significant difference between the 2 machine learning algorithms tested at the default settings with recall and precision values obtained. These findings demonstrate the promise of using a ML algorithm to assist in the performance of human tasks and behaviors, specifically the diagnosis of histopathologic images. These results have numerous potential uses in clinical medicine. ML algorithms have been successfully applied to diagnostic and prognostic endeavors in pathology,^23-28 dermatology,^29-31 ophthalmology,³² cardiology,³³ and radiology.^34-36

Pathologists often use additional tests, such as special staining of tissues or molecular tests, to assist with accurate classification of tumors. ML platforms offer the potential of an additional tool for pathologists to use along with human microscopic interpretation.^37,38 In addition, the number of pathologists in the US is dramatically decreasing, and many other countries have marked physician shortages, especially in fields of specialized training such as pathology.^39-42 These models could readily assist physicians in underserved countries and impact shortages of pathologists elsewhere by providing more specific diagnoses in an expedited manner.⁴³

Finally, although we have explored the application of these platforms in common cancer scenarios, great potential exists to use similar techniques in the detection of other conditions. These include the potential for classification and risk assessment of precancerous lesions, infectious processes in tissue (eg, detection of tuberculosis or malaria),^24,44 inflammatory conditions (eg, arthritis subtypes, gout),⁴⁵ blood disorders (eg, abnormal blood cell morphology),⁴⁶ and many others. The potential of these technologies to improve health care delivery to veteran patients seems to be limited only by the imagination of the user.⁴⁷

Regarding the limited effectiveness in determining the presence or absence of mutations in KRAS for colon adenocarcinoma, it is mentioned that currently pathologists rely on complex molecular tests to detect the mutations at the DNA level.²¹ It is possible that the use of more extensive training data sets may improve recall and precision in cases such as these and warrants further study. Our experiments were limited to the stipulations placed by the free trial software agreements; no costs were expended to use the algorithms, though an Apple computer was required.

Conclusion

We have demonstrated the successful application of 2 readily available ML platforms in providing diagnostic guidance in differentiation between common cancer conditions in veteran patient populations. Although both platforms performed very well with no statistically significant differences in results, some distinctions are worth noting. Apple Create ML can be used on local computers but is limited to an Apple operating system. Google AutoML is not platform-specific but runs only via Google Cloud with associated computational fees. Using these readily available models, we demonstrated the vast potential of AI in diagnostic pathology. The application of AI to clinical medicine remains in the very early stages. The VA is uniquely poised to provide leadership as AI technologies will continue to dramatically change the future of health care, both in veteran and nonveteran patients nationwide.

Acknowledgments

The authors thank Paul Borkowski for his constructive criticism and proofreading of this manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital.

Artificial intelligence (AI), first described in 1956, encompasses the field of computer science in which machines are trained to learn from experience. The term was popularized by the 1956 Dartmouth College Summer Research Project on Artificial Intelligence.¹ The field of AI is rapidly growing and has the potential to affect many aspects of our lives. The emerging importance of AI is demonstrated by a February 2019 executive order that launched the American AI Initiative, allocating resources and funding for AI development.² The executive order stresses the potential impact of AI in the health care field, including its potential utility to diagnose disease. Federal agencies were directed to invest in AI research and development to promote rapid breakthroughs in AI technology that may impact multiple areas of society.

Machine learning (ML), a subset of AI, was defined in 1959 by Arthur Samuel and is achieved by employing mathematic models to compute sample data sets.³ Originating from statistical linear models, neural networks were conceived to accomplish these tasks.⁴ These pioneering scientific achievements led to recent developments of deep neural networks. These models are developed to recognize patterns and achieve complex computational tasks within a matter of minutes, often far exceeding human ability.⁵ ML can increase efficiency with decreased computation time, high precision, and recall when compared with that of human decision making.⁶

ML has the potential for numerous applications in the health care field.^7-9 One promising application is in the field of anatomic pathology. ML allows representative images to be used to train a computer to recognize patterns from labeled photographs. Based on a set of images selected to represent a specific tissue or disease process, the computer can be trained to evaluate and recognize new and unique images from patients and render a diagnosis.¹⁰ Prior to modern ML models, users would have to import many thousands of training images to produce algorithms that could recognize patterns with high accuracy. Modern ML algorithms allow for a model known as transfer learning, such that far fewer images are required for training.^11-13

Two novel ML platforms available for public use are offered through Google (Mountain View, CA) and Apple (Cupertino, CA).^14,15 They each offer a user-friendly interface with minimal experience required in computer science. Google AutoML uses ML via cloud services to store and retrieve data with ease. No coding knowledge is required. The Apple Create ML Module provides computer-based ML, requiring only a few lines of code.

The Veterans Health Administration (VHA) is the largest single health care system in the US, and nearly 50 000 cancer cases are diagnosed at the VHA annually.¹⁶ Cancers of the lung and colon are among the most common sources of invasive cancer and are the 2 most common causes of cancer deaths in America.¹⁶ We have previously reported using Apple ML in detecting non-small cell lung cancers (NSCLCs), including adenocarcinomas and squamous cell carcinomas (SCCs); and colon cancers with accuracy.^17,18 In the present study, we expand on these findings by comparing Apple and Google ML platforms in a variety of common pathologic scenarios in veteran patients. Using limited training data, both programs are compared for precision and recall in differentiating conditions involving lung and colon pathology.

In the first 4 experiments, we evaluated the ability of the platforms to differentiate normal lung tissue from cancerous lung tissue, to distinguish lung adenocarcinoma from SCC, and to differentiate colon adenocarcinoma from normal colon tissue. Next, cases of colon adenocarcinoma were assessed to determine whether the presence or absence of the KRAS proto-oncogene could be determined histologically using the AI platforms. KRAS is found in a variety of cancers, including about 40% of colon adenocarcinomas.¹⁹ For colon cancers, the presence or absence of the mutation in KRAS has important implications for patients as it determines whether the tumor will respond to specific chemotherapy agents.²⁰ The presence of the KRAS gene is currently determined by complex molecular testing of tumor tissue.²¹ However, we assessed the potential of ML to determine whether the mutation is present by computerized morphologic analysis alone. Our last experiment examined the ability of the Apple and Google platforms to differentiate between adenocarcinomas of lung origin vs colon origin. This has potential utility in determining the site of origin of metastatic carcinoma.²²

Methods

Fifty cases of lung SCC, 50 cases of lung adenocarcinoma, and 50 cases of colon adenocarcinoma were randomly retrieved from our molecular database. Twenty-five colon adenocarcinoma cases were positive for mutation in KRAS, while 25 cases were negative for mutation in KRAS. Seven hundred fifty total images of lung tissue (250 benign lung tissue, 250 lung adenocarcinomas, and 250 lung SCCs) and 500 total images of colon tissue (250 benign colon tissue and 250 colon adenocarcinoma) were obtained using a Leica Microscope MC190 HD Camera (Wetzlar, Germany) connected to an Olympus BX41 microscope (Center Valley, PA) and the Leica Acquire 9072 software for Apple computers. All the images were captured at a resolution of 1024 x 768 pixels using a 60x dry objective. Lung tissue images were captured and saved on a 2012 Apple MacBook Pro computer, and colon images were captured and saved on a 2011 Apple iMac computer. Both computers were running macOS v10.13.

Creating Image Classifier Models Using Apple Create ML

Apple Create ML is a suite of products that use various tools to create and train custom ML models on Apple computers.¹⁵ The suite contains many features, including image classification to train a ML model to classify images, natural language processing to classify natural language text, and tabular data to train models that deal with labeling information or estimating new quantities. We used Create ML Image Classification to create image classifier models for our project (Appendix A).

Creating ML Modules Using Google Cloud AutoML Vision Beta

Google Cloud AutoML is a suite of machine learning products, including AutoML Vision, AutoML Natural Language and AutoML Translation.¹⁴ All Cloud AutoML machine learning products were in beta version at the time of experimentation. We used Cloud AutoML Vision beta to create ML modules for our project. Unlike Apple Create ML, which is run on a local Apple computer, the Google Cloud AutoML is run online using a Google Cloud account. There are no minimum specifications requirements for the local computer since it is using the cloud-based architecture (Appendix B).

Experiment 1

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect and subclassify NSCLC based on the histopathologic images. We created 3 classes of images (250 images each): benign lung tissue, lung adenocarcinoma, and lung SCC.

Experiment 2

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between normal lung tissue and NSCLC histopathologic images with 50/50 mixture of lung adenocarcinoma and lung SCC. We created 2 classes of images (250 images each): benign lung tissue and lung NSCLC.

Experiment 3

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and lung SCC histopathologic images. We created 2 classes of images (250 images each): adenocarcinoma and SCC.

Experiment 4

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect colon cancer histopathologic images regardless of mutation in KRAS status. We created 2 classes of images (250 images each): benign colon tissue and colon adenocarcinoma.

Experiment 5

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between colon adenocarcinoma with mutations in KRAS and colon adenocarcinoma without the mutation in KRAS histopathologic images. We created 2 classes of images (125 images each): colon adenocarcinoma cases with mutation in KRAS and colon adenocarcinoma cases without the mutation in KRAS.

Experiment 6

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and colon adenocarcinoma histopathologic images. We created 2 classes of images (250 images each): colon adenocarcinoma lung adenocarcinoma.

Results

Twelve machine learning models were created in 6 experiments using the Apple Create ML and the Google AutoML (Table). To investigate recall and precision differences between the Apple and the Google ML algorithms, we performed 2-tailed distribution, paired t tests. No statistically significant differences were found (P = .52 for recall and .60 for precision).

Overall, each model performed well in distinguishing between normal and neoplastic tissue for both lung and colon cancers. In subclassifying NSCLC into adenocarcinoma and SCC, the models were shown to have high levels of precision and recall. The models also were successful in distinguishing between lung and colonic origin of adenocarcinoma (Figures 1-4). However, both systems had trouble discerning colon adenocarcinoma with mutations in KRAS from adenocarcinoma without mutations in KRAS.

Discussion

Image classifier models using ML algorithms hold a promising future to revolutionize the health care field. ML products, such as those modules offered by Apple and Google, are easy to use and have a simple graphic user interface to allow individuals to train models to perform humanlike tasks in real time. In our experiments, we compared multiple algorithms to determine their ability to differentiate and subclassify histopathologic images with high precision and recall using common scenarios in treating veteran patients.

Analysis of the results revealed high precision and recall values illustrating the models’ ability to differentiate and detect benign lung tissue from lung SCC and lung adenocarcinoma in ML model 1, benign lung from NSCLC carcinoma in ML model 2, and benign colon from colonic adenocarcinoma in ML model 4. In ML model 3 and 6, both ML algorithms performed at a high level to differentiate lung SCC from lung adenocarcinoma and lung adenocarcinoma from colonic adenocarcinoma, respectively. Of note, ML model 5 had the lowest precision and recall values across both algorithms demonstrating the models’ limited utility in predicting molecular profiles, such as mutations in KRAS as tested here. This is not surprising as pathologists currently require complex molecular tests to detect mutations in KRAS reliably in colon cancer.

Both modules require minimal programming experience and are easy to use. In our comparison, we demonstrated critical distinguishing characteristics that differentiate the 2 products.

Apple Create ML image classifier is available for use on local Mac computers that use Xcode version 10 and macOS 10.14 or later, with just 3 lines of code required to perform computations. Although this product is limited to Apple computers, it is free to use, and images are stored on the computer hard drive. Of unique significance on the Apple system platform, images can be augmented to alter their appearance to enhance model training. For example, imported images can be cropped, rotated, blurred, and flipped, in order to optimize the model’s training abilities to recognize test images and perform pattern recognition. This feature is not as readily available on the Google platform. Apple Create ML Image classifier’s default training set consists of 75% of total imported images with 5% of the total images being randomly used as a validation set. The remaining 20% of images comprise the testing set. The module’s computational analysis to train the model is achieved in about 2 minutes on average. The score threshold is set at 50% and cannot be manipulated for each image class as in Google AutoML Vision.

Google AutoML Vision is open and can be accessed from many devices. It stores images on remote Google servers but requires computing fees after a $300 credit for 12 months. On AutoML Vision, random 80% of the total images are used in the training set, 10% are used in the validation set, and 10% are used in the testing set. It is important to highlight the different percentages used in the default settings on the respective modules. The time to train the Google AutoML Vision with default computational power is longer on average than Apple Create ML, with about 8 minutes required to train the machine learning module. However, it is possible to choose more computational power for an additional fee and decrease module training time. The user will receive e-mail alerts when the computer time begins and is completed. The computation time is calculated by subtracting the time of the initial e-mail from the final e-mail.

Based on our calculations, we determined there was no significant difference between the 2 machine learning algorithms tested at the default settings with recall and precision values obtained. These findings demonstrate the promise of using a ML algorithm to assist in the performance of human tasks and behaviors, specifically the diagnosis of histopathologic images. These results have numerous potential uses in clinical medicine. ML algorithms have been successfully applied to diagnostic and prognostic endeavors in pathology,^23-28 dermatology,^29-31 ophthalmology,³² cardiology,³³ and radiology.^34-36

Pathologists often use additional tests, such as special staining of tissues or molecular tests, to assist with accurate classification of tumors. ML platforms offer the potential of an additional tool for pathologists to use along with human microscopic interpretation.^37,38 In addition, the number of pathologists in the US is dramatically decreasing, and many other countries have marked physician shortages, especially in fields of specialized training such as pathology.^39-42 These models could readily assist physicians in underserved countries and impact shortages of pathologists elsewhere by providing more specific diagnoses in an expedited manner.⁴³

Finally, although we have explored the application of these platforms in common cancer scenarios, great potential exists to use similar techniques in the detection of other conditions. These include the potential for classification and risk assessment of precancerous lesions, infectious processes in tissue (eg, detection of tuberculosis or malaria),^24,44 inflammatory conditions (eg, arthritis subtypes, gout),⁴⁵ blood disorders (eg, abnormal blood cell morphology),⁴⁶ and many others. The potential of these technologies to improve health care delivery to veteran patients seems to be limited only by the imagination of the user.⁴⁷

Regarding the limited effectiveness in determining the presence or absence of mutations in KRAS for colon adenocarcinoma, it is mentioned that currently pathologists rely on complex molecular tests to detect the mutations at the DNA level.²¹ It is possible that the use of more extensive training data sets may improve recall and precision in cases such as these and warrants further study. Our experiments were limited to the stipulations placed by the free trial software agreements; no costs were expended to use the algorithms, though an Apple computer was required.

Conclusion

We have demonstrated the successful application of 2 readily available ML platforms in providing diagnostic guidance in differentiation between common cancer conditions in veteran patient populations. Although both platforms performed very well with no statistically significant differences in results, some distinctions are worth noting. Apple Create ML can be used on local computers but is limited to an Apple operating system. Google AutoML is not platform-specific but runs only via Google Cloud with associated computational fees. Using these readily available models, we demonstrated the vast potential of AI in diagnostic pathology. The application of AI to clinical medicine remains in the very early stages. The VA is uniquely poised to provide leadership as AI technologies will continue to dramatically change the future of health care, both in veteran and nonveteran patients nationwide.

Acknowledgments

The authors thank Paul Borkowski for his constructive criticism and proofreading of this manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital.

Artificial intelligence (AI), first described in 1956, encompasses the field of computer science in which machines are trained to learn from experience. The term was popularized by the 1956 Dartmouth College Summer Research Project on Artificial Intelligence.¹ The field of AI is rapidly growing and has the potential to affect many aspects of our lives. The emerging importance of AI is demonstrated by a February 2019 executive order that launched the American AI Initiative, allocating resources and funding for AI development.² The executive order stresses the potential impact of AI in the health care field, including its potential utility to diagnose disease. Federal agencies were directed to invest in AI research and development to promote rapid breakthroughs in AI technology that may impact multiple areas of society.

Machine learning (ML), a subset of AI, was defined in 1959 by Arthur Samuel and is achieved by employing mathematic models to compute sample data sets.³ Originating from statistical linear models, neural networks were conceived to accomplish these tasks.⁴ These pioneering scientific achievements led to recent developments of deep neural networks. These models are developed to recognize patterns and achieve complex computational tasks within a matter of minutes, often far exceeding human ability.⁵ ML can increase efficiency with decreased computation time, high precision, and recall when compared with that of human decision making.⁶

ML has the potential for numerous applications in the health care field.^7-9 One promising application is in the field of anatomic pathology. ML allows representative images to be used to train a computer to recognize patterns from labeled photographs. Based on a set of images selected to represent a specific tissue or disease process, the computer can be trained to evaluate and recognize new and unique images from patients and render a diagnosis.¹⁰ Prior to modern ML models, users would have to import many thousands of training images to produce algorithms that could recognize patterns with high accuracy. Modern ML algorithms allow for a model known as transfer learning, such that far fewer images are required for training.^11-13

Two novel ML platforms available for public use are offered through Google (Mountain View, CA) and Apple (Cupertino, CA).^14,15 They each offer a user-friendly interface with minimal experience required in computer science. Google AutoML uses ML via cloud services to store and retrieve data with ease. No coding knowledge is required. The Apple Create ML Module provides computer-based ML, requiring only a few lines of code.

The Veterans Health Administration (VHA) is the largest single health care system in the US, and nearly 50 000 cancer cases are diagnosed at the VHA annually.¹⁶ Cancers of the lung and colon are among the most common sources of invasive cancer and are the 2 most common causes of cancer deaths in America.¹⁶ We have previously reported using Apple ML in detecting non-small cell lung cancers (NSCLCs), including adenocarcinomas and squamous cell carcinomas (SCCs); and colon cancers with accuracy.^17,18 In the present study, we expand on these findings by comparing Apple and Google ML platforms in a variety of common pathologic scenarios in veteran patients. Using limited training data, both programs are compared for precision and recall in differentiating conditions involving lung and colon pathology.

In the first 4 experiments, we evaluated the ability of the platforms to differentiate normal lung tissue from cancerous lung tissue, to distinguish lung adenocarcinoma from SCC, and to differentiate colon adenocarcinoma from normal colon tissue. Next, cases of colon adenocarcinoma were assessed to determine whether the presence or absence of the KRAS proto-oncogene could be determined histologically using the AI platforms. KRAS is found in a variety of cancers, including about 40% of colon adenocarcinomas.¹⁹ For colon cancers, the presence or absence of the mutation in KRAS has important implications for patients as it determines whether the tumor will respond to specific chemotherapy agents.²⁰ The presence of the KRAS gene is currently determined by complex molecular testing of tumor tissue.²¹ However, we assessed the potential of ML to determine whether the mutation is present by computerized morphologic analysis alone. Our last experiment examined the ability of the Apple and Google platforms to differentiate between adenocarcinomas of lung origin vs colon origin. This has potential utility in determining the site of origin of metastatic carcinoma.²²

Methods

Fifty cases of lung SCC, 50 cases of lung adenocarcinoma, and 50 cases of colon adenocarcinoma were randomly retrieved from our molecular database. Twenty-five colon adenocarcinoma cases were positive for mutation in KRAS, while 25 cases were negative for mutation in KRAS. Seven hundred fifty total images of lung tissue (250 benign lung tissue, 250 lung adenocarcinomas, and 250 lung SCCs) and 500 total images of colon tissue (250 benign colon tissue and 250 colon adenocarcinoma) were obtained using a Leica Microscope MC190 HD Camera (Wetzlar, Germany) connected to an Olympus BX41 microscope (Center Valley, PA) and the Leica Acquire 9072 software for Apple computers. All the images were captured at a resolution of 1024 x 768 pixels using a 60x dry objective. Lung tissue images were captured and saved on a 2012 Apple MacBook Pro computer, and colon images were captured and saved on a 2011 Apple iMac computer. Both computers were running macOS v10.13.

Creating Image Classifier Models Using Apple Create ML

Apple Create ML is a suite of products that use various tools to create and train custom ML models on Apple computers.¹⁵ The suite contains many features, including image classification to train a ML model to classify images, natural language processing to classify natural language text, and tabular data to train models that deal with labeling information or estimating new quantities. We used Create ML Image Classification to create image classifier models for our project (Appendix A).

Creating ML Modules Using Google Cloud AutoML Vision Beta

Google Cloud AutoML is a suite of machine learning products, including AutoML Vision, AutoML Natural Language and AutoML Translation.¹⁴ All Cloud AutoML machine learning products were in beta version at the time of experimentation. We used Cloud AutoML Vision beta to create ML modules for our project. Unlike Apple Create ML, which is run on a local Apple computer, the Google Cloud AutoML is run online using a Google Cloud account. There are no minimum specifications requirements for the local computer since it is using the cloud-based architecture (Appendix B).

Experiment 1

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect and subclassify NSCLC based on the histopathologic images. We created 3 classes of images (250 images each): benign lung tissue, lung adenocarcinoma, and lung SCC.

Experiment 2

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between normal lung tissue and NSCLC histopathologic images with 50/50 mixture of lung adenocarcinoma and lung SCC. We created 2 classes of images (250 images each): benign lung tissue and lung NSCLC.

Experiment 3

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and lung SCC histopathologic images. We created 2 classes of images (250 images each): adenocarcinoma and SCC.

Experiment 4

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to detect colon cancer histopathologic images regardless of mutation in KRAS status. We created 2 classes of images (250 images each): benign colon tissue and colon adenocarcinoma.

Experiment 5

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between colon adenocarcinoma with mutations in KRAS and colon adenocarcinoma without the mutation in KRAS histopathologic images. We created 2 classes of images (125 images each): colon adenocarcinoma cases with mutation in KRAS and colon adenocarcinoma cases without the mutation in KRAS.

Experiment 6

We compared Apple Create ML Image Classifier and Google AutoML Vision in their ability to differentiate between lung adenocarcinoma and colon adenocarcinoma histopathologic images. We created 2 classes of images (250 images each): colon adenocarcinoma lung adenocarcinoma.

Results

Twelve machine learning models were created in 6 experiments using the Apple Create ML and the Google AutoML (Table). To investigate recall and precision differences between the Apple and the Google ML algorithms, we performed 2-tailed distribution, paired t tests. No statistically significant differences were found (P = .52 for recall and .60 for precision).

Overall, each model performed well in distinguishing between normal and neoplastic tissue for both lung and colon cancers. In subclassifying NSCLC into adenocarcinoma and SCC, the models were shown to have high levels of precision and recall. The models also were successful in distinguishing between lung and colonic origin of adenocarcinoma (Figures 1-4). However, both systems had trouble discerning colon adenocarcinoma with mutations in KRAS from adenocarcinoma without mutations in KRAS.

Discussion

Image classifier models using ML algorithms hold a promising future to revolutionize the health care field. ML products, such as those modules offered by Apple and Google, are easy to use and have a simple graphic user interface to allow individuals to train models to perform humanlike tasks in real time. In our experiments, we compared multiple algorithms to determine their ability to differentiate and subclassify histopathologic images with high precision and recall using common scenarios in treating veteran patients.

Analysis of the results revealed high precision and recall values illustrating the models’ ability to differentiate and detect benign lung tissue from lung SCC and lung adenocarcinoma in ML model 1, benign lung from NSCLC carcinoma in ML model 2, and benign colon from colonic adenocarcinoma in ML model 4. In ML model 3 and 6, both ML algorithms performed at a high level to differentiate lung SCC from lung adenocarcinoma and lung adenocarcinoma from colonic adenocarcinoma, respectively. Of note, ML model 5 had the lowest precision and recall values across both algorithms demonstrating the models’ limited utility in predicting molecular profiles, such as mutations in KRAS as tested here. This is not surprising as pathologists currently require complex molecular tests to detect mutations in KRAS reliably in colon cancer.

Both modules require minimal programming experience and are easy to use. In our comparison, we demonstrated critical distinguishing characteristics that differentiate the 2 products.

Apple Create ML image classifier is available for use on local Mac computers that use Xcode version 10 and macOS 10.14 or later, with just 3 lines of code required to perform computations. Although this product is limited to Apple computers, it is free to use, and images are stored on the computer hard drive. Of unique significance on the Apple system platform, images can be augmented to alter their appearance to enhance model training. For example, imported images can be cropped, rotated, blurred, and flipped, in order to optimize the model’s training abilities to recognize test images and perform pattern recognition. This feature is not as readily available on the Google platform. Apple Create ML Image classifier’s default training set consists of 75% of total imported images with 5% of the total images being randomly used as a validation set. The remaining 20% of images comprise the testing set. The module’s computational analysis to train the model is achieved in about 2 minutes on average. The score threshold is set at 50% and cannot be manipulated for each image class as in Google AutoML Vision.

Google AutoML Vision is open and can be accessed from many devices. It stores images on remote Google servers but requires computing fees after a $300 credit for 12 months. On AutoML Vision, random 80% of the total images are used in the training set, 10% are used in the validation set, and 10% are used in the testing set. It is important to highlight the different percentages used in the default settings on the respective modules. The time to train the Google AutoML Vision with default computational power is longer on average than Apple Create ML, with about 8 minutes required to train the machine learning module. However, it is possible to choose more computational power for an additional fee and decrease module training time. The user will receive e-mail alerts when the computer time begins and is completed. The computation time is calculated by subtracting the time of the initial e-mail from the final e-mail.

Based on our calculations, we determined there was no significant difference between the 2 machine learning algorithms tested at the default settings with recall and precision values obtained. These findings demonstrate the promise of using a ML algorithm to assist in the performance of human tasks and behaviors, specifically the diagnosis of histopathologic images. These results have numerous potential uses in clinical medicine. ML algorithms have been successfully applied to diagnostic and prognostic endeavors in pathology,^23-28 dermatology,^29-31 ophthalmology,³² cardiology,³³ and radiology.^34-36

Pathologists often use additional tests, such as special staining of tissues or molecular tests, to assist with accurate classification of tumors. ML platforms offer the potential of an additional tool for pathologists to use along with human microscopic interpretation.^37,38 In addition, the number of pathologists in the US is dramatically decreasing, and many other countries have marked physician shortages, especially in fields of specialized training such as pathology.^39-42 These models could readily assist physicians in underserved countries and impact shortages of pathologists elsewhere by providing more specific diagnoses in an expedited manner.⁴³

Finally, although we have explored the application of these platforms in common cancer scenarios, great potential exists to use similar techniques in the detection of other conditions. These include the potential for classification and risk assessment of precancerous lesions, infectious processes in tissue (eg, detection of tuberculosis or malaria),^24,44 inflammatory conditions (eg, arthritis subtypes, gout),⁴⁵ blood disorders (eg, abnormal blood cell morphology),⁴⁶ and many others. The potential of these technologies to improve health care delivery to veteran patients seems to be limited only by the imagination of the user.⁴⁷

Regarding the limited effectiveness in determining the presence or absence of mutations in KRAS for colon adenocarcinoma, it is mentioned that currently pathologists rely on complex molecular tests to detect the mutations at the DNA level.²¹ It is possible that the use of more extensive training data sets may improve recall and precision in cases such as these and warrants further study. Our experiments were limited to the stipulations placed by the free trial software agreements; no costs were expended to use the algorithms, though an Apple computer was required.

Conclusion

We have demonstrated the successful application of 2 readily available ML platforms in providing diagnostic guidance in differentiation between common cancer conditions in veteran patient populations. Although both platforms performed very well with no statistically significant differences in results, some distinctions are worth noting. Apple Create ML can be used on local computers but is limited to an Apple operating system. Google AutoML is not platform-specific but runs only via Google Cloud with associated computational fees. Using these readily available models, we demonstrated the vast potential of AI in diagnostic pathology. The application of AI to clinical medicine remains in the very early stages. The VA is uniquely poised to provide leadership as AI technologies will continue to dramatically change the future of health care, both in veteran and nonveteran patients nationwide.

Acknowledgments

The authors thank Paul Borkowski for his constructive criticism and proofreading of this manuscript. This material is the result of work supported with resources and the use of facilities at the James A. Haley Veterans’ Hospital.

Physician burnout has been labeled a public health crisis by the Harvard School of Public Health and other institutions.¹ A 2018 Physician’s Foundation survey found that 78% of physicians had symptoms of burnout,² which result from chronic workplace stress and include feeling depleted of energy or exhausted, mentally distanced from or cynical about one’s job, and problems getting one’s job done successfully.³ Among ObGyns, almost half (46%) report burnout.⁴ One-third of ObGyns responded on a recent Medscape Burnout Report that the computerization of practice is contributing to their burnout, and 54% said too many bureaucratic tasks, including charting, were adding to their burnout.⁵

Inefficient electronic medical records (EMRs) have been implicated as one reason for burnout, with improvements in efficiency cited as one of several potential resolutions to the problem. About 96% of hospitals have adopted EMRs today, compared with only 9% in 2008,⁶ and many physicians report recognizing value in the technology. For instance, 60% of participants in Stanford Medicine’s 2018 National Physician Poll said EMRs had led to improved patient care. At the same time, however, about as many (59%) said EMRs needed a “complete overhaul” and that the systems had detracted from their professional satisfaction (54%) as well as from their clinical effectiveness (49%).⁷

With this roundtable, we explore the concerns with hours spent on the EMR with several experts, and whether it is a problem that has been contributing to burnout among staff at their institutions. In addition, are there solutions that their institutions have implemented that they can share to help to cope with the problem?
 

OBG Management: ObGyns report that the computerization of practice and too many bureaucratic tasks, including charting, are contributing to burnout. Do you see this problem at your institution?

John J. Dougherty, MD, MBA: Yes, absolutely. There is not a day that goes by that I don’t hear about or experience “Epic Fails.” (We use Epic’s EMR product at our institution.) Too many clicks are needed to navigate even the simplest tasks—finding notes or results, documenting visits, and billing for services are all unnecessarily complex. In addition, we are being held accountable for achieving a long and growing list of “metrics” measures, education projects (HealthStream), and productivity goals. When do we have time to treat patients? And it is not just practicing physicians and clinicians. Our resident physicians spend an inordinate amount of time in front of the computer documenting, placing orders, and transferring patients using a system with a very inefficient user interface, to say the least.

Megan L. Evans, MD, MPH: I absolutely agree. Over the years, my institution has created a conglomerate of EMRs, requiring physicians across the hospital to be fluent in a multitude of systems. For example, you finish your clinic notes in one system, sign off on discharge summaries in another, and complete your operative notes in an entirely different system. As busy attendings, it is hard to keep ahead of all of these tasks, especially when the systems do not talk to one another. Fortunately, my hospital is changing our EMR to a single system within the next year. Until then, however, we will work in this piecemeal system.

Mark Woodland, MS, MD: EMR and computerization of medicine is the number 1 issue relating to dissatisfaction by ObGyn providers in our institution. Providers are earnest in their attempt to be compliant with EMR requirements, but the reality is that they are dealing with an automated system that does not have realistic expectations for management of results, follow-up tasks, and patient communications for a human provider. The actual charting, ordering of tests and consults, and communication between providers has been enhanced. However, the “in-basket” of tasks to be accomplished are extraordinary and much of it relies on the provider, which requires an inordinate amount of time. Additionally, while other members of the medical staff are stationary at a desk, physicians and other providers are not. They are mobile between inpatient units, labor and delivery, operating rooms, and emergency rooms. Time management does not always allow for providers to access computers from all of these areas to facilitate their managing the expectations of the EMR. This requires providers to access the EMR at off hours, extending their workload. Finally, the EMR is neither personal nor friendly. It is not designed with the clinician in mind, and it is not fun or engaging for a provider.

Continue to: OBG Management: What solutions have been instituted...

OBG Management: What solutions have been instituted in order to help physicians with data entry into the EMR?

Dr. Dougherty: When our institution compared EMR offerings, EMR companies put their best collective marketing feet forward. The general notion, at least with the Epic EMR, was that “you can customize Epic to your liking.” It did not take long for a bunch of motivated Epic users to create “smart” stuff (lists, phrases, and texts) in an effort to customize workflows and create fancy-looking electronic notes. Shortly thereafter, it was obvious that, as an institution, our reporting efforts kept coming up short—our reports lacked accuracy and meaning. Everyone was documenting in different ways and in different areas. Considering that reports are currently generated using (mostly) discrete data entries (data placed in specific fields within the EMR), it became obvious that our data entry paradigm needed to change. Therefore, standardization became the leading buzzword. Our institution recently initiated a project aimed at standardizing our workflows and documentation habits. In addition, we have incorporated a third-party information exchange product into our health system data aggregation and analysis workflow. Much more needs to be done, but it is a start.

Dr. Evans: At my institution, as a group, we have created templates for routine procedures and visits that also auto populate billing codes. I know that some departments have used scribes. From the hospital side, there has been improved access to the EMR from home. Some of my colleagues like this feature; however, others, like myself, believe this contributes to some of our burnout. I like to leave work at work. Having the ability to continue working at home is not a solution in my mind.

Dr. Woodland: At our institution, we have engaged our chaperones and medical assistants to help facilitate completion of the medical records during the office visit. Providers work with their assistants to accommodate documentation of history and physical findings while also listening to the provider as they are speaking in order to document patient care plans and orders. This saves the clinicians time in reviewing and editing the record as well as making sure the appropriate care plan is instituted. Our EMR provider recently has begun experimenting with personalization of color themes as well as pictures as part of the interface. Having said this, I still ask, “Why have medical professionals allowed non–clinical agencies and information technology groups to run this show?” It is also inconceivable to me that this unfunded mandate—that has increased cost, decreased clinical efficiency, and decreased clinician satisfaction—has not been addressed by national and international medical communities.

OBG Management: What changes do you feel your EMR system needs to undergo?

Dr. Woodland: I feel that we need to appropriately manage expectations of the EMR and the institution with relation to EMR and providers. By this I mean that we need to make the EMR more user-friendly and appropriate for different clinicians as well as patients. We also need to manage expectations of our patients. In a digital age where immediate contact is the norm, we need to address the issue that the EMR is not social media but rather a communication tool for routine contact and information transmission. Emergencies are not typically addressed well through the EMR platform; they are better handled with a more appropriate communication interface.

Dr. Dougherty: I feel that the biggest change needed is a competent, simple, and standard user-interface. Our old charting methods were great on a number of levels. For instance, if I wanted to add an order, I flipped to the ”Orders” tab and entered an order. If I needed to document a note, I flipped to the “Notes” tab and started writing, etc. Obviously, manual charting had its downsides—like trying to decipher handwriting art! EMRs could easily adopt the stuff that worked from our old methods of documentation, while leveraging the advantages that computerized workflows can bring to practitioners, including efficient transfer of records, meaningful reporting, simple electronic ordering, and interprofessional communication portals.

Dr. Evans: Our systems need to better communicate with one another. I am in an academic practice, and I should be able to see labs, consultant notes, imaging, all in one spot to improve efficiency and ease with patient visits. Minimizing clicks would be helpful as well. I try to write as much as I can while in the room with a patient to avoid after-hours note writing, but it takes away from my interaction with each patient.

Continue to: OBG Management: With an aim toward alleviating burnout...

OBG Management: With an aim toward alleviating burnout, are there any tips you can offer your colleagues on interfacing with the EMR?

Dr. Evans: When I first started as a new attending, it would take me hours to finish my notes, partly because of the level of detail I would write in my history of present illness (HPI) and assessment and plan. One great piece of advice I received was to be satisfied with good notes, not perfect notes. I worked to consolidate my thoughts and use preconstructed phrases/paragraphs on common problems I saw. This saved time to focus on other aspects of my academic job.

Dr. Dougherty: We need to refocus on the patient first, and mold our systems to meet that priority. Much too often, we have our backs to the patients or spend too much time interfacing with our EMR systems, and our patients are not happy about it (as many surveys have demonstrated). More importantly, a renewed focus on patient care, not EMR care, would allow our practitioners to do what they signed up for—treating patients. In the meantime, I would suggest that practitioners stay away from EMR gimmicks and go back to old-style documentation practices (like those established by the Centers for Medicare and Medicaid Services in 1997 and 1998), and ask the IT folks to help with molding the EMR systems to meet your own standards, not the standards established by EMR companies. I am also very hopeful that the consumer will drive most of the health care-related data collection in the near future, thereby marginalizing the current generation of EMR systems.

Dr. Woodland: I would add that providers need to manage the EMR and not let the EMR manage them. Set up task reminders at point times to handle results and communications from the EMR and set up time in your schedule where you can facilitate meeting these tasks. When providers are out on vacation, make sure to have an out-of-office reminder built into their EMR so that patients and others know timing of potential responses. Try to make the EMR as enjoyable as possible and focus on the good points of the EMR, such as legibility, order verification, safety, and documentation.
 

OBG Management: Do you feel that the EMR has led to improved patient care?

Dr. Evans: Yes and no. Yes, in that it can be much easier to follow a patient’s health care history from other provider notes or prior surgeries. Information is searchable and legible. If an EMR is built correctly, it can save time for providers, through smart phrases and templates, and it can help providers with proper billing codes and documentation requirements. No, in that it can take away from important patient interaction. We are required to see more patients in less time all while using, at times, a cumbersome EMR system.

Dr. Woodland: This is a tricky question because the EMR has both positive and negative attributes. Certainly, the legibility and order verification has improved, but the ease of accessing information in the EMR has changed. Additionally, there has been a drastic increase in provider dissatisfaction that has not been addressed. Provider dissatisfaction can lead to problems in patient care. If there was a clear-cut increased value for the cost, I do not think the EMR would be such a huge focus of negative attention. Providers need to take back control of their EMR and their profession so that they can utilize the EMR as the tool it was supposed to be and not the dissatisfier that it has become.

Dr. Dougherty: I do not believe patient care has been improved by EMR systems, for all of the reasons we have discussed, and then some. But there is an enormous amount of potential, if we get the interface between humans and EMR systems right!

Physician burnout has been labeled a public health crisis by the Harvard School of Public Health and other institutions.¹ A 2018 Physician’s Foundation survey found that 78% of physicians had symptoms of burnout,² which result from chronic workplace stress and include feeling depleted of energy or exhausted, mentally distanced from or cynical about one’s job, and problems getting one’s job done successfully.³ Among ObGyns, almost half (46%) report burnout.⁴ One-third of ObGyns responded on a recent Medscape Burnout Report that the computerization of practice is contributing to their burnout, and 54% said too many bureaucratic tasks, including charting, were adding to their burnout.⁵

Inefficient electronic medical records (EMRs) have been implicated as one reason for burnout, with improvements in efficiency cited as one of several potential resolutions to the problem. About 96% of hospitals have adopted EMRs today, compared with only 9% in 2008,⁶ and many physicians report recognizing value in the technology. For instance, 60% of participants in Stanford Medicine’s 2018 National Physician Poll said EMRs had led to improved patient care. At the same time, however, about as many (59%) said EMRs needed a “complete overhaul” and that the systems had detracted from their professional satisfaction (54%) as well as from their clinical effectiveness (49%).⁷

With this roundtable, we explore the concerns with hours spent on the EMR with several experts, and whether it is a problem that has been contributing to burnout among staff at their institutions. In addition, are there solutions that their institutions have implemented that they can share to help to cope with the problem?
 

OBG Management: ObGyns report that the computerization of practice and too many bureaucratic tasks, including charting, are contributing to burnout. Do you see this problem at your institution?

John J. Dougherty, MD, MBA: Yes, absolutely. There is not a day that goes by that I don’t hear about or experience “Epic Fails.” (We use Epic’s EMR product at our institution.) Too many clicks are needed to navigate even the simplest tasks—finding notes or results, documenting visits, and billing for services are all unnecessarily complex. In addition, we are being held accountable for achieving a long and growing list of “metrics” measures, education projects (HealthStream), and productivity goals. When do we have time to treat patients? And it is not just practicing physicians and clinicians. Our resident physicians spend an inordinate amount of time in front of the computer documenting, placing orders, and transferring patients using a system with a very inefficient user interface, to say the least.

Megan L. Evans, MD, MPH: I absolutely agree. Over the years, my institution has created a conglomerate of EMRs, requiring physicians across the hospital to be fluent in a multitude of systems. For example, you finish your clinic notes in one system, sign off on discharge summaries in another, and complete your operative notes in an entirely different system. As busy attendings, it is hard to keep ahead of all of these tasks, especially when the systems do not talk to one another. Fortunately, my hospital is changing our EMR to a single system within the next year. Until then, however, we will work in this piecemeal system.

Mark Woodland, MS, MD: EMR and computerization of medicine is the number 1 issue relating to dissatisfaction by ObGyn providers in our institution. Providers are earnest in their attempt to be compliant with EMR requirements, but the reality is that they are dealing with an automated system that does not have realistic expectations for management of results, follow-up tasks, and patient communications for a human provider. The actual charting, ordering of tests and consults, and communication between providers has been enhanced. However, the “in-basket” of tasks to be accomplished are extraordinary and much of it relies on the provider, which requires an inordinate amount of time. Additionally, while other members of the medical staff are stationary at a desk, physicians and other providers are not. They are mobile between inpatient units, labor and delivery, operating rooms, and emergency rooms. Time management does not always allow for providers to access computers from all of these areas to facilitate their managing the expectations of the EMR. This requires providers to access the EMR at off hours, extending their workload. Finally, the EMR is neither personal nor friendly. It is not designed with the clinician in mind, and it is not fun or engaging for a provider.

Continue to: OBG Management: What solutions have been instituted...

OBG Management: What solutions have been instituted in order to help physicians with data entry into the EMR?

Dr. Dougherty: When our institution compared EMR offerings, EMR companies put their best collective marketing feet forward. The general notion, at least with the Epic EMR, was that “you can customize Epic to your liking.” It did not take long for a bunch of motivated Epic users to create “smart” stuff (lists, phrases, and texts) in an effort to customize workflows and create fancy-looking electronic notes. Shortly thereafter, it was obvious that, as an institution, our reporting efforts kept coming up short—our reports lacked accuracy and meaning. Everyone was documenting in different ways and in different areas. Considering that reports are currently generated using (mostly) discrete data entries (data placed in specific fields within the EMR), it became obvious that our data entry paradigm needed to change. Therefore, standardization became the leading buzzword. Our institution recently initiated a project aimed at standardizing our workflows and documentation habits. In addition, we have incorporated a third-party information exchange product into our health system data aggregation and analysis workflow. Much more needs to be done, but it is a start.

Dr. Evans: At my institution, as a group, we have created templates for routine procedures and visits that also auto populate billing codes. I know that some departments have used scribes. From the hospital side, there has been improved access to the EMR from home. Some of my colleagues like this feature; however, others, like myself, believe this contributes to some of our burnout. I like to leave work at work. Having the ability to continue working at home is not a solution in my mind.

Dr. Woodland: At our institution, we have engaged our chaperones and medical assistants to help facilitate completion of the medical records during the office visit. Providers work with their assistants to accommodate documentation of history and physical findings while also listening to the provider as they are speaking in order to document patient care plans and orders. This saves the clinicians time in reviewing and editing the record as well as making sure the appropriate care plan is instituted. Our EMR provider recently has begun experimenting with personalization of color themes as well as pictures as part of the interface. Having said this, I still ask, “Why have medical professionals allowed non–clinical agencies and information technology groups to run this show?” It is also inconceivable to me that this unfunded mandate—that has increased cost, decreased clinical efficiency, and decreased clinician satisfaction—has not been addressed by national and international medical communities.

OBG Management: What changes do you feel your EMR system needs to undergo?

Dr. Woodland: I feel that we need to appropriately manage expectations of the EMR and the institution with relation to EMR and providers. By this I mean that we need to make the EMR more user-friendly and appropriate for different clinicians as well as patients. We also need to manage expectations of our patients. In a digital age where immediate contact is the norm, we need to address the issue that the EMR is not social media but rather a communication tool for routine contact and information transmission. Emergencies are not typically addressed well through the EMR platform; they are better handled with a more appropriate communication interface.

Dr. Dougherty: I feel that the biggest change needed is a competent, simple, and standard user-interface. Our old charting methods were great on a number of levels. For instance, if I wanted to add an order, I flipped to the ”Orders” tab and entered an order. If I needed to document a note, I flipped to the “Notes” tab and started writing, etc. Obviously, manual charting had its downsides—like trying to decipher handwriting art! EMRs could easily adopt the stuff that worked from our old methods of documentation, while leveraging the advantages that computerized workflows can bring to practitioners, including efficient transfer of records, meaningful reporting, simple electronic ordering, and interprofessional communication portals.

Dr. Evans: Our systems need to better communicate with one another. I am in an academic practice, and I should be able to see labs, consultant notes, imaging, all in one spot to improve efficiency and ease with patient visits. Minimizing clicks would be helpful as well. I try to write as much as I can while in the room with a patient to avoid after-hours note writing, but it takes away from my interaction with each patient.

Continue to: OBG Management: With an aim toward alleviating burnout...

OBG Management: With an aim toward alleviating burnout, are there any tips you can offer your colleagues on interfacing with the EMR?

Dr. Evans: When I first started as a new attending, it would take me hours to finish my notes, partly because of the level of detail I would write in my history of present illness (HPI) and assessment and plan. One great piece of advice I received was to be satisfied with good notes, not perfect notes. I worked to consolidate my thoughts and use preconstructed phrases/paragraphs on common problems I saw. This saved time to focus on other aspects of my academic job.

Dr. Dougherty: We need to refocus on the patient first, and mold our systems to meet that priority. Much too often, we have our backs to the patients or spend too much time interfacing with our EMR systems, and our patients are not happy about it (as many surveys have demonstrated). More importantly, a renewed focus on patient care, not EMR care, would allow our practitioners to do what they signed up for—treating patients. In the meantime, I would suggest that practitioners stay away from EMR gimmicks and go back to old-style documentation practices (like those established by the Centers for Medicare and Medicaid Services in 1997 and 1998), and ask the IT folks to help with molding the EMR systems to meet your own standards, not the standards established by EMR companies. I am also very hopeful that the consumer will drive most of the health care-related data collection in the near future, thereby marginalizing the current generation of EMR systems.

Dr. Woodland: I would add that providers need to manage the EMR and not let the EMR manage them. Set up task reminders at point times to handle results and communications from the EMR and set up time in your schedule where you can facilitate meeting these tasks. When providers are out on vacation, make sure to have an out-of-office reminder built into their EMR so that patients and others know timing of potential responses. Try to make the EMR as enjoyable as possible and focus on the good points of the EMR, such as legibility, order verification, safety, and documentation.
 

OBG Management: Do you feel that the EMR has led to improved patient care?

Dr. Evans: Yes and no. Yes, in that it can be much easier to follow a patient’s health care history from other provider notes or prior surgeries. Information is searchable and legible. If an EMR is built correctly, it can save time for providers, through smart phrases and templates, and it can help providers with proper billing codes and documentation requirements. No, in that it can take away from important patient interaction. We are required to see more patients in less time all while using, at times, a cumbersome EMR system.

Dr. Woodland: This is a tricky question because the EMR has both positive and negative attributes. Certainly, the legibility and order verification has improved, but the ease of accessing information in the EMR has changed. Additionally, there has been a drastic increase in provider dissatisfaction that has not been addressed. Provider dissatisfaction can lead to problems in patient care. If there was a clear-cut increased value for the cost, I do not think the EMR would be such a huge focus of negative attention. Providers need to take back control of their EMR and their profession so that they can utilize the EMR as the tool it was supposed to be and not the dissatisfier that it has become.

Dr. Dougherty: I do not believe patient care has been improved by EMR systems, for all of the reasons we have discussed, and then some. But there is an enormous amount of potential, if we get the interface between humans and EMR systems right!

Physician burnout has been labeled a public health crisis by the Harvard School of Public Health and other institutions.¹ A 2018 Physician’s Foundation survey found that 78% of physicians had symptoms of burnout,² which result from chronic workplace stress and include feeling depleted of energy or exhausted, mentally distanced from or cynical about one’s job, and problems getting one’s job done successfully.³ Among ObGyns, almost half (46%) report burnout.⁴ One-third of ObGyns responded on a recent Medscape Burnout Report that the computerization of practice is contributing to their burnout, and 54% said too many bureaucratic tasks, including charting, were adding to their burnout.⁵

Inefficient electronic medical records (EMRs) have been implicated as one reason for burnout, with improvements in efficiency cited as one of several potential resolutions to the problem. About 96% of hospitals have adopted EMRs today, compared with only 9% in 2008,⁶ and many physicians report recognizing value in the technology. For instance, 60% of participants in Stanford Medicine’s 2018 National Physician Poll said EMRs had led to improved patient care. At the same time, however, about as many (59%) said EMRs needed a “complete overhaul” and that the systems had detracted from their professional satisfaction (54%) as well as from their clinical effectiveness (49%).⁷

With this roundtable, we explore the concerns with hours spent on the EMR with several experts, and whether it is a problem that has been contributing to burnout among staff at their institutions. In addition, are there solutions that their institutions have implemented that they can share to help to cope with the problem?
 

OBG Management: ObGyns report that the computerization of practice and too many bureaucratic tasks, including charting, are contributing to burnout. Do you see this problem at your institution?

John J. Dougherty, MD, MBA: Yes, absolutely. There is not a day that goes by that I don’t hear about or experience “Epic Fails.” (We use Epic’s EMR product at our institution.) Too many clicks are needed to navigate even the simplest tasks—finding notes or results, documenting visits, and billing for services are all unnecessarily complex. In addition, we are being held accountable for achieving a long and growing list of “metrics” measures, education projects (HealthStream), and productivity goals. When do we have time to treat patients? And it is not just practicing physicians and clinicians. Our resident physicians spend an inordinate amount of time in front of the computer documenting, placing orders, and transferring patients using a system with a very inefficient user interface, to say the least.

Megan L. Evans, MD, MPH: I absolutely agree. Over the years, my institution has created a conglomerate of EMRs, requiring physicians across the hospital to be fluent in a multitude of systems. For example, you finish your clinic notes in one system, sign off on discharge summaries in another, and complete your operative notes in an entirely different system. As busy attendings, it is hard to keep ahead of all of these tasks, especially when the systems do not talk to one another. Fortunately, my hospital is changing our EMR to a single system within the next year. Until then, however, we will work in this piecemeal system.

Mark Woodland, MS, MD: EMR and computerization of medicine is the number 1 issue relating to dissatisfaction by ObGyn providers in our institution. Providers are earnest in their attempt to be compliant with EMR requirements, but the reality is that they are dealing with an automated system that does not have realistic expectations for management of results, follow-up tasks, and patient communications for a human provider. The actual charting, ordering of tests and consults, and communication between providers has been enhanced. However, the “in-basket” of tasks to be accomplished are extraordinary and much of it relies on the provider, which requires an inordinate amount of time. Additionally, while other members of the medical staff are stationary at a desk, physicians and other providers are not. They are mobile between inpatient units, labor and delivery, operating rooms, and emergency rooms. Time management does not always allow for providers to access computers from all of these areas to facilitate their managing the expectations of the EMR. This requires providers to access the EMR at off hours, extending their workload. Finally, the EMR is neither personal nor friendly. It is not designed with the clinician in mind, and it is not fun or engaging for a provider.

Continue to: OBG Management: What solutions have been instituted...

OBG Management: What solutions have been instituted in order to help physicians with data entry into the EMR?

Dr. Dougherty: When our institution compared EMR offerings, EMR companies put their best collective marketing feet forward. The general notion, at least with the Epic EMR, was that “you can customize Epic to your liking.” It did not take long for a bunch of motivated Epic users to create “smart” stuff (lists, phrases, and texts) in an effort to customize workflows and create fancy-looking electronic notes. Shortly thereafter, it was obvious that, as an institution, our reporting efforts kept coming up short—our reports lacked accuracy and meaning. Everyone was documenting in different ways and in different areas. Considering that reports are currently generated using (mostly) discrete data entries (data placed in specific fields within the EMR), it became obvious that our data entry paradigm needed to change. Therefore, standardization became the leading buzzword. Our institution recently initiated a project aimed at standardizing our workflows and documentation habits. In addition, we have incorporated a third-party information exchange product into our health system data aggregation and analysis workflow. Much more needs to be done, but it is a start.

Dr. Evans: At my institution, as a group, we have created templates for routine procedures and visits that also auto populate billing codes. I know that some departments have used scribes. From the hospital side, there has been improved access to the EMR from home. Some of my colleagues like this feature; however, others, like myself, believe this contributes to some of our burnout. I like to leave work at work. Having the ability to continue working at home is not a solution in my mind.

Dr. Woodland: At our institution, we have engaged our chaperones and medical assistants to help facilitate completion of the medical records during the office visit. Providers work with their assistants to accommodate documentation of history and physical findings while also listening to the provider as they are speaking in order to document patient care plans and orders. This saves the clinicians time in reviewing and editing the record as well as making sure the appropriate care plan is instituted. Our EMR provider recently has begun experimenting with personalization of color themes as well as pictures as part of the interface. Having said this, I still ask, “Why have medical professionals allowed non–clinical agencies and information technology groups to run this show?” It is also inconceivable to me that this unfunded mandate—that has increased cost, decreased clinical efficiency, and decreased clinician satisfaction—has not been addressed by national and international medical communities.

OBG Management: What changes do you feel your EMR system needs to undergo?

Dr. Woodland: I feel that we need to appropriately manage expectations of the EMR and the institution with relation to EMR and providers. By this I mean that we need to make the EMR more user-friendly and appropriate for different clinicians as well as patients. We also need to manage expectations of our patients. In a digital age where immediate contact is the norm, we need to address the issue that the EMR is not social media but rather a communication tool for routine contact and information transmission. Emergencies are not typically addressed well through the EMR platform; they are better handled with a more appropriate communication interface.

Dr. Dougherty: I feel that the biggest change needed is a competent, simple, and standard user-interface. Our old charting methods were great on a number of levels. For instance, if I wanted to add an order, I flipped to the ”Orders” tab and entered an order. If I needed to document a note, I flipped to the “Notes” tab and started writing, etc. Obviously, manual charting had its downsides—like trying to decipher handwriting art! EMRs could easily adopt the stuff that worked from our old methods of documentation, while leveraging the advantages that computerized workflows can bring to practitioners, including efficient transfer of records, meaningful reporting, simple electronic ordering, and interprofessional communication portals.

Dr. Evans: Our systems need to better communicate with one another. I am in an academic practice, and I should be able to see labs, consultant notes, imaging, all in one spot to improve efficiency and ease with patient visits. Minimizing clicks would be helpful as well. I try to write as much as I can while in the room with a patient to avoid after-hours note writing, but it takes away from my interaction with each patient.

Continue to: OBG Management: With an aim toward alleviating burnout...

OBG Management: With an aim toward alleviating burnout, are there any tips you can offer your colleagues on interfacing with the EMR?

Dr. Evans: When I first started as a new attending, it would take me hours to finish my notes, partly because of the level of detail I would write in my history of present illness (HPI) and assessment and plan. One great piece of advice I received was to be satisfied with good notes, not perfect notes. I worked to consolidate my thoughts and use preconstructed phrases/paragraphs on common problems I saw. This saved time to focus on other aspects of my academic job.

Dr. Dougherty: We need to refocus on the patient first, and mold our systems to meet that priority. Much too often, we have our backs to the patients or spend too much time interfacing with our EMR systems, and our patients are not happy about it (as many surveys have demonstrated). More importantly, a renewed focus on patient care, not EMR care, would allow our practitioners to do what they signed up for—treating patients. In the meantime, I would suggest that practitioners stay away from EMR gimmicks and go back to old-style documentation practices (like those established by the Centers for Medicare and Medicaid Services in 1997 and 1998), and ask the IT folks to help with molding the EMR systems to meet your own standards, not the standards established by EMR companies. I am also very hopeful that the consumer will drive most of the health care-related data collection in the near future, thereby marginalizing the current generation of EMR systems.

Dr. Woodland: I would add that providers need to manage the EMR and not let the EMR manage them. Set up task reminders at point times to handle results and communications from the EMR and set up time in your schedule where you can facilitate meeting these tasks. When providers are out on vacation, make sure to have an out-of-office reminder built into their EMR so that patients and others know timing of potential responses. Try to make the EMR as enjoyable as possible and focus on the good points of the EMR, such as legibility, order verification, safety, and documentation.
 

OBG Management: Do you feel that the EMR has led to improved patient care?

Dr. Evans: Yes and no. Yes, in that it can be much easier to follow a patient’s health care history from other provider notes or prior surgeries. Information is searchable and legible. If an EMR is built correctly, it can save time for providers, through smart phrases and templates, and it can help providers with proper billing codes and documentation requirements. No, in that it can take away from important patient interaction. We are required to see more patients in less time all while using, at times, a cumbersome EMR system.

Dr. Woodland: This is a tricky question because the EMR has both positive and negative attributes. Certainly, the legibility and order verification has improved, but the ease of accessing information in the EMR has changed. Additionally, there has been a drastic increase in provider dissatisfaction that has not been addressed. Provider dissatisfaction can lead to problems in patient care. If there was a clear-cut increased value for the cost, I do not think the EMR would be such a huge focus of negative attention. Providers need to take back control of their EMR and their profession so that they can utilize the EMR as the tool it was supposed to be and not the dissatisfier that it has become.

Dr. Dougherty: I do not believe patient care has been improved by EMR systems, for all of the reasons we have discussed, and then some. But there is an enormous amount of potential, if we get the interface between humans and EMR systems right!

When first responders arrive at a scene where illicit drugs may be present, they could be at risk of dangerous exposure. They might inhale drugs; they can have contact through mucous membranes or through needlesticks.

A major concern is exposure to fentanyl or its analogues, which can lead to symptoms, including rapid onset of life-threatening respiratory depression. The exception is skin contact, which is not expected to have toxic effects if the visible contamination is removed promptly.

To help EMS providers and other responders protect themselves, the National Institute for Occupational Safety and Health (NIOSH) has released a new virtual toolkit with videos, infographics, and postcards based on NIOSH safety recommendations.

The resources highlight how best to assess the scene for hazards that may indicate the presence of illicit drugs and what to do—for example, use soap and water, not hand sanitizer (it doesn’t remove illicit drugs and may increase exposure), and don’t eat, drink, smoke, or use the bathroom in the affected area. The infographics also show how to decontaminate and prevent “take-home exposure” to protect responders’ families. The guidelines extend to procedures for protecting working dogs exposed to the drugs.

NIOSH notes that it has no occupational exposure data on fentanyl or its analogues for emergency responders. The recommendations are based on the reported toxicity and the chemical and physical properties of fentanyl and its analogues, NIOSH guidance for similar chemicals, recommendations from previous NIOSH health hazard evaluation reports, and “the basic principles of industrial hygiene.” As new research becomes available, NIOSH says, the recommendations will be updated.

The toolkit resources are shareable and available for disseminating via print, social media, text, and more. The kit is accessible at https://www.cdc.gov/niosh/topics/fentanyl/toolkit.html.

When first responders arrive at a scene where illicit drugs may be present, they could be at risk of dangerous exposure. They might inhale drugs; they can have contact through mucous membranes or through needlesticks.

A major concern is exposure to fentanyl or its analogues, which can lead to symptoms, including rapid onset of life-threatening respiratory depression. The exception is skin contact, which is not expected to have toxic effects if the visible contamination is removed promptly.

To help EMS providers and other responders protect themselves, the National Institute for Occupational Safety and Health (NIOSH) has released a new virtual toolkit with videos, infographics, and postcards based on NIOSH safety recommendations.

The resources highlight how best to assess the scene for hazards that may indicate the presence of illicit drugs and what to do—for example, use soap and water, not hand sanitizer (it doesn’t remove illicit drugs and may increase exposure), and don’t eat, drink, smoke, or use the bathroom in the affected area. The infographics also show how to decontaminate and prevent “take-home exposure” to protect responders’ families. The guidelines extend to procedures for protecting working dogs exposed to the drugs.

NIOSH notes that it has no occupational exposure data on fentanyl or its analogues for emergency responders. The recommendations are based on the reported toxicity and the chemical and physical properties of fentanyl and its analogues, NIOSH guidance for similar chemicals, recommendations from previous NIOSH health hazard evaluation reports, and “the basic principles of industrial hygiene.” As new research becomes available, NIOSH says, the recommendations will be updated.

The toolkit resources are shareable and available for disseminating via print, social media, text, and more. The kit is accessible at https://www.cdc.gov/niosh/topics/fentanyl/toolkit.html.

When first responders arrive at a scene where illicit drugs may be present, they could be at risk of dangerous exposure. They might inhale drugs; they can have contact through mucous membranes or through needlesticks.

A major concern is exposure to fentanyl or its analogues, which can lead to symptoms, including rapid onset of life-threatening respiratory depression. The exception is skin contact, which is not expected to have toxic effects if the visible contamination is removed promptly.

To help EMS providers and other responders protect themselves, the National Institute for Occupational Safety and Health (NIOSH) has released a new virtual toolkit with videos, infographics, and postcards based on NIOSH safety recommendations.

The resources highlight how best to assess the scene for hazards that may indicate the presence of illicit drugs and what to do—for example, use soap and water, not hand sanitizer (it doesn’t remove illicit drugs and may increase exposure), and don’t eat, drink, smoke, or use the bathroom in the affected area. The infographics also show how to decontaminate and prevent “take-home exposure” to protect responders’ families. The guidelines extend to procedures for protecting working dogs exposed to the drugs.

NIOSH notes that it has no occupational exposure data on fentanyl or its analogues for emergency responders. The recommendations are based on the reported toxicity and the chemical and physical properties of fentanyl and its analogues, NIOSH guidance for similar chemicals, recommendations from previous NIOSH health hazard evaluation reports, and “the basic principles of industrial hygiene.” As new research becomes available, NIOSH says, the recommendations will be updated.

The toolkit resources are shareable and available for disseminating via print, social media, text, and more. The kit is accessible at https://www.cdc.gov/niosh/topics/fentanyl/toolkit.html.

The updated smoking policy goes into effect for employees, patients, visitors, volunteers, contractors, and vendors, whether they smoke cigarettes, cigars, pipes, or even electronic and vaping devices, and whenever they are on the grounds of VA health care facilities, including parking areas.

The new policy comes after the VA reviewed research on second- and thirdhand smoke and best practices in the health care industry. “There is no risk-free level of exposure to tobacco smoke,” the VA’s Smokefree website says. Overwhelming evidence shows exposure to secondhand smoke has significant medical risks. Moreover, a growing body of evidence shows exposure to thirdhand smoke (residual nicotine and other chemicals left on indoor surfaces) also is a health hazard. The residue is thought to react with indoor pollutants to create a toxic mix that clings long after smoking has stopped and cannot be eliminated by opening windows, or using fans, or other means of clearing rooms.

“We are not alone in recognizing the importance of creating a smoke-free campus,” said VA Secretary Robert Wilkie. He notes that as of 2014, 4000 health care facilities and 4 national health care systems in the US have implemented smoke-free grounds.

National Association of Government employees will begin implementing the policy as of October 1, and have until January 1, 2020, to fully comply. Smoking shelters will be closed, although each facility will independently determine the disposition of smoking areas and shelters.

The new policy does not mean anyone has to quit smoking but to encourage quitting, the VA offers resources, including www.publichealth.va.gov/smoking/quit/index.asp. More tips and tools are available at the Smokefree Veteran website: https://veterans.smokefree.gov. SmokefreeVET is a text-messaging program (https://veterans.smokefree.gov/tools-tips-vet/smokefreevet) that provides 24/7 support to help veterans quit for good. Employees can contact their facility for resources.

The policies are available at https://www.va.gov/health/smokefree.

The updated smoking policy goes into effect for employees, patients, visitors, volunteers, contractors, and vendors, whether they smoke cigarettes, cigars, pipes, or even electronic and vaping devices, and whenever they are on the grounds of VA health care facilities, including parking areas.

The new policy comes after the VA reviewed research on second- and thirdhand smoke and best practices in the health care industry. “There is no risk-free level of exposure to tobacco smoke,” the VA’s Smokefree website says. Overwhelming evidence shows exposure to secondhand smoke has significant medical risks. Moreover, a growing body of evidence shows exposure to thirdhand smoke (residual nicotine and other chemicals left on indoor surfaces) also is a health hazard. The residue is thought to react with indoor pollutants to create a toxic mix that clings long after smoking has stopped and cannot be eliminated by opening windows, or using fans, or other means of clearing rooms.

“We are not alone in recognizing the importance of creating a smoke-free campus,” said VA Secretary Robert Wilkie. He notes that as of 2014, 4000 health care facilities and 4 national health care systems in the US have implemented smoke-free grounds.

National Association of Government employees will begin implementing the policy as of October 1, and have until January 1, 2020, to fully comply. Smoking shelters will be closed, although each facility will independently determine the disposition of smoking areas and shelters.

The new policy does not mean anyone has to quit smoking but to encourage quitting, the VA offers resources, including www.publichealth.va.gov/smoking/quit/index.asp. More tips and tools are available at the Smokefree Veteran website: https://veterans.smokefree.gov. SmokefreeVET is a text-messaging program (https://veterans.smokefree.gov/tools-tips-vet/smokefreevet) that provides 24/7 support to help veterans quit for good. Employees can contact their facility for resources.

The policies are available at https://www.va.gov/health/smokefree.

The updated smoking policy goes into effect for employees, patients, visitors, volunteers, contractors, and vendors, whether they smoke cigarettes, cigars, pipes, or even electronic and vaping devices, and whenever they are on the grounds of VA health care facilities, including parking areas.

The new policy comes after the VA reviewed research on second- and thirdhand smoke and best practices in the health care industry. “There is no risk-free level of exposure to tobacco smoke,” the VA’s Smokefree website says. Overwhelming evidence shows exposure to secondhand smoke has significant medical risks. Moreover, a growing body of evidence shows exposure to thirdhand smoke (residual nicotine and other chemicals left on indoor surfaces) also is a health hazard. The residue is thought to react with indoor pollutants to create a toxic mix that clings long after smoking has stopped and cannot be eliminated by opening windows, or using fans, or other means of clearing rooms.

“We are not alone in recognizing the importance of creating a smoke-free campus,” said VA Secretary Robert Wilkie. He notes that as of 2014, 4000 health care facilities and 4 national health care systems in the US have implemented smoke-free grounds.

National Association of Government employees will begin implementing the policy as of October 1, and have until January 1, 2020, to fully comply. Smoking shelters will be closed, although each facility will independently determine the disposition of smoking areas and shelters.

The new policy does not mean anyone has to quit smoking but to encourage quitting, the VA offers resources, including www.publichealth.va.gov/smoking/quit/index.asp. More tips and tools are available at the Smokefree Veteran website: https://veterans.smokefree.gov. SmokefreeVET is a text-messaging program (https://veterans.smokefree.gov/tools-tips-vet/smokefreevet) that provides 24/7 support to help veterans quit for good. Employees can contact their facility for resources.

The policies are available at https://www.va.gov/health/smokefree.

Heparin is the anticoagulant of choice when a rapid anticoagulant is indicated: Onset of action is immediate when administered IV as a bolus.¹ The major anticoagulant effect of heparin is mediated by heparin/antithrombin (AT) interaction. Heparin/AT inactivates factor IIa (thrombin) and factors Xa, IXa, XIa, and XIIa. Heparin is approved for multiple indications, such as venous thromboembolism (VTE) treatment and prophylaxis of medical and surgical patients; stroke prevention in atrial fibrillation (AF); acute coronary syndrome (ACS); vascular and cardiac surgeries; and various interventional procedures (eg, diagnostic angiography and percutaneous coronary intervention [PCI]). It also is used as an anticoagulant in blood transfusions, extracorporeal circulation, and for maintaining patency of central vascular access devices (CVADs).

About 60% of the crude heparin used to manufacture heparin in the US originates in China, derived from porcine mucosa. African swine fever, a contagious virus with no cure, has eliminated about 25% to 35% of China’s pig population, or about 150 million pigs. In July 2019, members of the US House of Representatives Committee on Energy and Commerce sent a letter to the US Food and Drug Administration asking for details on the potential impact of African swine fever on the supply of heparin.²

The US Department of Veterans Affairs (VA) heath care system is currently experiencing a shortage of heparin vials and syringes. It is unclear when resolution of this shortage will occur as it could resolve within several weeks or as late as January 2020.³ Although vials and syringes are the current products that are affected, it is possible the shortage may eventually include IV heparin bags as well.

Since the foremost objective of VA health care providers is to provide timely access to medications for veterans, strategies to conserve unfractionated heparin (UfH) must be used since it is a first-line therapy where few evidence-based alternatives exist. Conservation strategies may include drug rationing, therapeutic substitution, and compounding of needed products using the limited stock available in the pharmacy.⁴ It is important that all staff are educated on facility strategies in order to be familiar with alternatives and limit the potential for near misses, adverse events, and provider frustration.

In shortage situations, the VA-Pharmacy Benefits Management (PBM) defers decisions regarding drug preservation, processes to shift to viable alternatives, and the best practice for safe transitions to local facilities and their subject matter experts.⁵ At the VA Tennessee Valley Healthcare System, a 1A, tertiary, dual campus health care system, a pharmacy task force has formed to track drug shortages impacting the facility’s efficiencies and budgets. This group communicates with the Pharmacy and Therapeutics committee about potential risks to patient care and develops shortage briefs (following an SBAR [situation, background, assessment, recommendation] design) generally authored and championed by at least 1 clinical pharmacy specialist and supervising physicians who are field experts. Prior to dissemination, the SBAR undergoes a rapid peer-review process.

To date, VA PBM has not issued specific guidance on how pharmacists should proceed in case of a shortage. However, we recommend strategies that may be considered for implementation during a potential UfH shortage. For example, pharmacists can use therapeutic alternatives for which best available evidence suggests no disadvantage.⁴ The Table lists alternative agents according to indication and patient-specific considerations that may preclude use. Existing UfH products may also be used for drug compounding (eg, use current stock to provide an indicated aliquot) to meet the need of prioritized patients.⁴ In addition, we suggest prioritizing current UfH/heparinized saline for use for the following groups of patients⁴:

• Emergent/urgent cardiac surgery^1,6;
• Hemodialysis patients^1,7-9 for which the low-molecular-weight heparin (LMWH) dalteparin is deemed inappropriate or the patient is not monitored in the intensive care unit for regional citrate administration;
• VTE prophylaxis for patients with epidurals or chest tubes for which urgent invasive management may occur, recent cardiac or neurosurgery, or for patients with a creatine clearance < 15 mL/min or receiving hemodialysis^10-12;
• Vascular surgery (eg, limb ischemia) and interventions (eg, carotid stenting, endarterectomy)^13,14;
• Mesenteric ischemia (venous thrombosis) with a potential to proceed to laparotomy¹⁵;
• Critically ill patients with arterial lines for which normal saline is deemed inappropriate for line flushing¹⁶;
• Electrophysiology procedures (eg, AF ablation)¹⁷; and
• Contraindication to use of a long-acting alternative listed in the table or a medical necessity exists for using a rapidly reversible agent. Examples for this category include but are not limited to recent gastrointestinal bleeding, central nervous system lesion, and select neurologic diagnoses (eg, cerebral venous sinus thrombosis with hemorrhage, thrombus in vertebral basilar system or anterior circulation, intraparenchymal hemorrhage plus mechanical valve, medium to large cardioembolic stroke with intracardiac thrombus).

Conclusion

The UfH drug shortage represents a significant threat to public health and is a major challenge for US health care systems, including the Veterans Health Administration. Overreliance on a predominant source of crude heparin has affected multiple UfH manufacturers and products. Current alternatives to UfH include low-molecular-weight heparins, IV direct thrombin inhibitors, and SC fondaparinux, with selection supported by guidelines or evolving literature. However, the shortage has the potential to expand to other injectables, such as dalteparin and enoxaparin, and severely limit care for veterans. It is vital that clinicians rapidly address the current shortage by creating a plan to develop efficient and equitable access to UfH, continue to assess supply and update stakeholders, and select evidence-based alternatives while maintaining focus on efficacy and safety.

Acknowledgments

The authors thank Ashley Yost, PharmD, for her coordination of the multidisciplinary task force assigned to efficiently manage the heparin drug shortage. This material is the result of work supported with resources and the use of facilities at the VA Tennessee Valley Healthcare System in Nashville, Tennessee.

Heparin is the anticoagulant of choice when a rapid anticoagulant is indicated: Onset of action is immediate when administered IV as a bolus.¹ The major anticoagulant effect of heparin is mediated by heparin/antithrombin (AT) interaction. Heparin/AT inactivates factor IIa (thrombin) and factors Xa, IXa, XIa, and XIIa. Heparin is approved for multiple indications, such as venous thromboembolism (VTE) treatment and prophylaxis of medical and surgical patients; stroke prevention in atrial fibrillation (AF); acute coronary syndrome (ACS); vascular and cardiac surgeries; and various interventional procedures (eg, diagnostic angiography and percutaneous coronary intervention [PCI]). It also is used as an anticoagulant in blood transfusions, extracorporeal circulation, and for maintaining patency of central vascular access devices (CVADs).

About 60% of the crude heparin used to manufacture heparin in the US originates in China, derived from porcine mucosa. African swine fever, a contagious virus with no cure, has eliminated about 25% to 35% of China’s pig population, or about 150 million pigs. In July 2019, members of the US House of Representatives Committee on Energy and Commerce sent a letter to the US Food and Drug Administration asking for details on the potential impact of African swine fever on the supply of heparin.²

The US Department of Veterans Affairs (VA) heath care system is currently experiencing a shortage of heparin vials and syringes. It is unclear when resolution of this shortage will occur as it could resolve within several weeks or as late as January 2020.³ Although vials and syringes are the current products that are affected, it is possible the shortage may eventually include IV heparin bags as well.

Since the foremost objective of VA health care providers is to provide timely access to medications for veterans, strategies to conserve unfractionated heparin (UfH) must be used since it is a first-line therapy where few evidence-based alternatives exist. Conservation strategies may include drug rationing, therapeutic substitution, and compounding of needed products using the limited stock available in the pharmacy.⁴ It is important that all staff are educated on facility strategies in order to be familiar with alternatives and limit the potential for near misses, adverse events, and provider frustration.

In shortage situations, the VA-Pharmacy Benefits Management (PBM) defers decisions regarding drug preservation, processes to shift to viable alternatives, and the best practice for safe transitions to local facilities and their subject matter experts.⁵ At the VA Tennessee Valley Healthcare System, a 1A, tertiary, dual campus health care system, a pharmacy task force has formed to track drug shortages impacting the facility’s efficiencies and budgets. This group communicates with the Pharmacy and Therapeutics committee about potential risks to patient care and develops shortage briefs (following an SBAR [situation, background, assessment, recommendation] design) generally authored and championed by at least 1 clinical pharmacy specialist and supervising physicians who are field experts. Prior to dissemination, the SBAR undergoes a rapid peer-review process.

To date, VA PBM has not issued specific guidance on how pharmacists should proceed in case of a shortage. However, we recommend strategies that may be considered for implementation during a potential UfH shortage. For example, pharmacists can use therapeutic alternatives for which best available evidence suggests no disadvantage.⁴ The Table lists alternative agents according to indication and patient-specific considerations that may preclude use. Existing UfH products may also be used for drug compounding (eg, use current stock to provide an indicated aliquot) to meet the need of prioritized patients.⁴ In addition, we suggest prioritizing current UfH/heparinized saline for use for the following groups of patients⁴:

• Emergent/urgent cardiac surgery^1,6;
• Hemodialysis patients^1,7-9 for which the low-molecular-weight heparin (LMWH) dalteparin is deemed inappropriate or the patient is not monitored in the intensive care unit for regional citrate administration;
• VTE prophylaxis for patients with epidurals or chest tubes for which urgent invasive management may occur, recent cardiac or neurosurgery, or for patients with a creatine clearance < 15 mL/min or receiving hemodialysis^10-12;
• Vascular surgery (eg, limb ischemia) and interventions (eg, carotid stenting, endarterectomy)^13,14;
• Mesenteric ischemia (venous thrombosis) with a potential to proceed to laparotomy¹⁵;
• Critically ill patients with arterial lines for which normal saline is deemed inappropriate for line flushing¹⁶;
• Electrophysiology procedures (eg, AF ablation)¹⁷; and
• Contraindication to use of a long-acting alternative listed in the table or a medical necessity exists for using a rapidly reversible agent. Examples for this category include but are not limited to recent gastrointestinal bleeding, central nervous system lesion, and select neurologic diagnoses (eg, cerebral venous sinus thrombosis with hemorrhage, thrombus in vertebral basilar system or anterior circulation, intraparenchymal hemorrhage plus mechanical valve, medium to large cardioembolic stroke with intracardiac thrombus).

Conclusion

The UfH drug shortage represents a significant threat to public health and is a major challenge for US health care systems, including the Veterans Health Administration. Overreliance on a predominant source of crude heparin has affected multiple UfH manufacturers and products. Current alternatives to UfH include low-molecular-weight heparins, IV direct thrombin inhibitors, and SC fondaparinux, with selection supported by guidelines or evolving literature. However, the shortage has the potential to expand to other injectables, such as dalteparin and enoxaparin, and severely limit care for veterans. It is vital that clinicians rapidly address the current shortage by creating a plan to develop efficient and equitable access to UfH, continue to assess supply and update stakeholders, and select evidence-based alternatives while maintaining focus on efficacy and safety.

Acknowledgments

The authors thank Ashley Yost, PharmD, for her coordination of the multidisciplinary task force assigned to efficiently manage the heparin drug shortage. This material is the result of work supported with resources and the use of facilities at the VA Tennessee Valley Healthcare System in Nashville, Tennessee.

Heparin is the anticoagulant of choice when a rapid anticoagulant is indicated: Onset of action is immediate when administered IV as a bolus.¹ The major anticoagulant effect of heparin is mediated by heparin/antithrombin (AT) interaction. Heparin/AT inactivates factor IIa (thrombin) and factors Xa, IXa, XIa, and XIIa. Heparin is approved for multiple indications, such as venous thromboembolism (VTE) treatment and prophylaxis of medical and surgical patients; stroke prevention in atrial fibrillation (AF); acute coronary syndrome (ACS); vascular and cardiac surgeries; and various interventional procedures (eg, diagnostic angiography and percutaneous coronary intervention [PCI]). It also is used as an anticoagulant in blood transfusions, extracorporeal circulation, and for maintaining patency of central vascular access devices (CVADs).

About 60% of the crude heparin used to manufacture heparin in the US originates in China, derived from porcine mucosa. African swine fever, a contagious virus with no cure, has eliminated about 25% to 35% of China’s pig population, or about 150 million pigs. In July 2019, members of the US House of Representatives Committee on Energy and Commerce sent a letter to the US Food and Drug Administration asking for details on the potential impact of African swine fever on the supply of heparin.²

The US Department of Veterans Affairs (VA) heath care system is currently experiencing a shortage of heparin vials and syringes. It is unclear when resolution of this shortage will occur as it could resolve within several weeks or as late as January 2020.³ Although vials and syringes are the current products that are affected, it is possible the shortage may eventually include IV heparin bags as well.

Since the foremost objective of VA health care providers is to provide timely access to medications for veterans, strategies to conserve unfractionated heparin (UfH) must be used since it is a first-line therapy where few evidence-based alternatives exist. Conservation strategies may include drug rationing, therapeutic substitution, and compounding of needed products using the limited stock available in the pharmacy.⁴ It is important that all staff are educated on facility strategies in order to be familiar with alternatives and limit the potential for near misses, adverse events, and provider frustration.

In shortage situations, the VA-Pharmacy Benefits Management (PBM) defers decisions regarding drug preservation, processes to shift to viable alternatives, and the best practice for safe transitions to local facilities and their subject matter experts.⁵ At the VA Tennessee Valley Healthcare System, a 1A, tertiary, dual campus health care system, a pharmacy task force has formed to track drug shortages impacting the facility’s efficiencies and budgets. This group communicates with the Pharmacy and Therapeutics committee about potential risks to patient care and develops shortage briefs (following an SBAR [situation, background, assessment, recommendation] design) generally authored and championed by at least 1 clinical pharmacy specialist and supervising physicians who are field experts. Prior to dissemination, the SBAR undergoes a rapid peer-review process.

To date, VA PBM has not issued specific guidance on how pharmacists should proceed in case of a shortage. However, we recommend strategies that may be considered for implementation during a potential UfH shortage. For example, pharmacists can use therapeutic alternatives for which best available evidence suggests no disadvantage.⁴ The Table lists alternative agents according to indication and patient-specific considerations that may preclude use. Existing UfH products may also be used for drug compounding (eg, use current stock to provide an indicated aliquot) to meet the need of prioritized patients.⁴ In addition, we suggest prioritizing current UfH/heparinized saline for use for the following groups of patients⁴:

• Emergent/urgent cardiac surgery^1,6;
• Hemodialysis patients^1,7-9 for which the low-molecular-weight heparin (LMWH) dalteparin is deemed inappropriate or the patient is not monitored in the intensive care unit for regional citrate administration;
• VTE prophylaxis for patients with epidurals or chest tubes for which urgent invasive management may occur, recent cardiac or neurosurgery, or for patients with a creatine clearance < 15 mL/min or receiving hemodialysis^10-12;
• Vascular surgery (eg, limb ischemia) and interventions (eg, carotid stenting, endarterectomy)^13,14;
• Mesenteric ischemia (venous thrombosis) with a potential to proceed to laparotomy¹⁵;
• Critically ill patients with arterial lines for which normal saline is deemed inappropriate for line flushing¹⁶;
• Electrophysiology procedures (eg, AF ablation)¹⁷; and
• Contraindication to use of a long-acting alternative listed in the table or a medical necessity exists for using a rapidly reversible agent. Examples for this category include but are not limited to recent gastrointestinal bleeding, central nervous system lesion, and select neurologic diagnoses (eg, cerebral venous sinus thrombosis with hemorrhage, thrombus in vertebral basilar system or anterior circulation, intraparenchymal hemorrhage plus mechanical valve, medium to large cardioembolic stroke with intracardiac thrombus).

Conclusion

The UfH drug shortage represents a significant threat to public health and is a major challenge for US health care systems, including the Veterans Health Administration. Overreliance on a predominant source of crude heparin has affected multiple UfH manufacturers and products. Current alternatives to UfH include low-molecular-weight heparins, IV direct thrombin inhibitors, and SC fondaparinux, with selection supported by guidelines or evolving literature. However, the shortage has the potential to expand to other injectables, such as dalteparin and enoxaparin, and severely limit care for veterans. It is vital that clinicians rapidly address the current shortage by creating a plan to develop efficient and equitable access to UfH, continue to assess supply and update stakeholders, and select evidence-based alternatives while maintaining focus on efficacy and safety.

Acknowledgments

The authors thank Ashley Yost, PharmD, for her coordination of the multidisciplinary task force assigned to efficiently manage the heparin drug shortage. This material is the result of work supported with resources and the use of facilities at the VA Tennessee Valley Healthcare System in Nashville, Tennessee.

CASE Patient has adverse effects from halted estrogen pills

JR twists her hands nervously as you step into the room. “They stopped my hormones,” she sighs as you pull up her lab results.

JR recently had been admitted to an inpatient cardiology unit for several days for a heart failure exacerbation. Her ankles are still swollen beneath her floral print skirt, but she is breathing much easier now. She is back at your primary care office, hoping to get clearance to restart her estrogen pills.

JR reports having mood swings and terrible nightmares while not taking her hormones, which she has been taking for more than 3 years. She hesitates before sharing, “One of the doctors kept asking me questions about my sex life that had nothing to do with my heart condition. I don’t want to go back there.”

Providing compassionate and comprehensive care to gender-nonconforming individuals is challenging for a multitude of reasons, from clinician ignorance to systemic discrimination. About 33% of transgender patients reported being harassed, denied care, or even being assaulted when seeking health care, while 23% reported avoiding going to the doctor altogether when sick or injured out of fear of discrimination.¹

Unfortunately, now, further increases to barriers to care may be put in place. In late May of this year, the Department of Health and Human Services (HHS) proposed new regulations that would reverse previous regulations granted through section 1557 of the Affordable Care Act (ACA)—the Health Care Rights Law—which affirmed the rights of gender nonbinary persons to medical care. Among the proposed changes is the elimination of protections against discrimination in health care based on gender identity.² The proposed regulation changes come on the heels of a federal court case, which seeks to declare that hospital systems may turn away patients based on gender identity.³

Unraveling rights afforded under the ACA

The Health Care Rights Law was passed under the ACA; it prohibits discrimination based on race, color, national origin, sex, age, and disability in health programs and activities receiving federal financial assistance. Multiple lower courts have supported that the rights of transgender individuals is included within these protections against discrimination on the basis of sex.⁴ These court rulings not only have ensured the ability of gender-diverse individuals to access care but also have enforced insurance coverage of therapies for gender dysphoria. It was only in 2014 that Medicaid began providing coverage for gender-affirming surgeries and eliminating language that such procedures were “experimental” or “cosmetic.” The 2016 passage of the ACA mandated that private insurance companies follow suit. Unfortunately, the recent proposed regulation changes to the Health Care Rights Law may spark a reversal from insurance companies as well. Such a setback would affect gender-diverse individuals’ hormone treatments as well as their ability to access a full spectrum of care within the health care system.

Continue to: ACOG urges nondiscriminatory practices...

ACOG urges nondiscriminatory practices

The proposed regulation changes to the Health Care Rights Law are from the Conscience and Religious Freedom Division of the HHS Office for Civil Rights, which was established in 2018 and has been advocating for the rights of health care providers to refuse to treat patients based on their own religious beliefs.⁵ We argue, however, that providing care to persons of varying backgrounds is not an assault on our individual liberties but rather a privilege as providers. As obstetrician-gynecologists, it may be easy to only consider cis-gendered women our responsibility. But our field also emphasizes individual empowerment above all else—we fight every day for our patients’ rights to contraception, fertility, pregnancy, parenthood, and sexual freedoms. Let us continue speaking up for the rights of all those who need gynecologic care, regardless of the pronouns they use.

“The American College of Obstetricians and Gynecologists urges health care providers to foster nondiscriminatory practices and policies to increase identification and to facilitate quality health care for transgender individuals, both in assisting with the transition if desired as well as providing long-term preventive health care.”⁶

We urge you to take action

• Reach out to your local representatives about protecting transgender health access
• Educate yourself on the unique needs of transgender individuals
• Read personal accounts
• Share your personal story
• Find referring providers near your practice

CASE Patient has adverse effects from halted estrogen pills

JR twists her hands nervously as you step into the room. “They stopped my hormones,” she sighs as you pull up her lab results.

JR recently had been admitted to an inpatient cardiology unit for several days for a heart failure exacerbation. Her ankles are still swollen beneath her floral print skirt, but she is breathing much easier now. She is back at your primary care office, hoping to get clearance to restart her estrogen pills.

JR reports having mood swings and terrible nightmares while not taking her hormones, which she has been taking for more than 3 years. She hesitates before sharing, “One of the doctors kept asking me questions about my sex life that had nothing to do with my heart condition. I don’t want to go back there.”

Providing compassionate and comprehensive care to gender-nonconforming individuals is challenging for a multitude of reasons, from clinician ignorance to systemic discrimination. About 33% of transgender patients reported being harassed, denied care, or even being assaulted when seeking health care, while 23% reported avoiding going to the doctor altogether when sick or injured out of fear of discrimination.¹

Unfortunately, now, further increases to barriers to care may be put in place. In late May of this year, the Department of Health and Human Services (HHS) proposed new regulations that would reverse previous regulations granted through section 1557 of the Affordable Care Act (ACA)—the Health Care Rights Law—which affirmed the rights of gender nonbinary persons to medical care. Among the proposed changes is the elimination of protections against discrimination in health care based on gender identity.² The proposed regulation changes come on the heels of a federal court case, which seeks to declare that hospital systems may turn away patients based on gender identity.³

Unraveling rights afforded under the ACA

The Health Care Rights Law was passed under the ACA; it prohibits discrimination based on race, color, national origin, sex, age, and disability in health programs and activities receiving federal financial assistance. Multiple lower courts have supported that the rights of transgender individuals is included within these protections against discrimination on the basis of sex.⁴ These court rulings not only have ensured the ability of gender-diverse individuals to access care but also have enforced insurance coverage of therapies for gender dysphoria. It was only in 2014 that Medicaid began providing coverage for gender-affirming surgeries and eliminating language that such procedures were “experimental” or “cosmetic.” The 2016 passage of the ACA mandated that private insurance companies follow suit. Unfortunately, the recent proposed regulation changes to the Health Care Rights Law may spark a reversal from insurance companies as well. Such a setback would affect gender-diverse individuals’ hormone treatments as well as their ability to access a full spectrum of care within the health care system.

Continue to: ACOG urges nondiscriminatory practices...

ACOG urges nondiscriminatory practices

The proposed regulation changes to the Health Care Rights Law are from the Conscience and Religious Freedom Division of the HHS Office for Civil Rights, which was established in 2018 and has been advocating for the rights of health care providers to refuse to treat patients based on their own religious beliefs.⁵ We argue, however, that providing care to persons of varying backgrounds is not an assault on our individual liberties but rather a privilege as providers. As obstetrician-gynecologists, it may be easy to only consider cis-gendered women our responsibility. But our field also emphasizes individual empowerment above all else—we fight every day for our patients’ rights to contraception, fertility, pregnancy, parenthood, and sexual freedoms. Let us continue speaking up for the rights of all those who need gynecologic care, regardless of the pronouns they use.

“The American College of Obstetricians and Gynecologists urges health care providers to foster nondiscriminatory practices and policies to increase identification and to facilitate quality health care for transgender individuals, both in assisting with the transition if desired as well as providing long-term preventive health care.”⁶

We urge you to take action

• Reach out to your local representatives about protecting transgender health access
• Educate yourself on the unique needs of transgender individuals
• Read personal accounts
• Share your personal story
• Find referring providers near your practice

CASE Patient has adverse effects from halted estrogen pills

JR twists her hands nervously as you step into the room. “They stopped my hormones,” she sighs as you pull up her lab results.

JR recently had been admitted to an inpatient cardiology unit for several days for a heart failure exacerbation. Her ankles are still swollen beneath her floral print skirt, but she is breathing much easier now. She is back at your primary care office, hoping to get clearance to restart her estrogen pills.

JR reports having mood swings and terrible nightmares while not taking her hormones, which she has been taking for more than 3 years. She hesitates before sharing, “One of the doctors kept asking me questions about my sex life that had nothing to do with my heart condition. I don’t want to go back there.”

Providing compassionate and comprehensive care to gender-nonconforming individuals is challenging for a multitude of reasons, from clinician ignorance to systemic discrimination. About 33% of transgender patients reported being harassed, denied care, or even being assaulted when seeking health care, while 23% reported avoiding going to the doctor altogether when sick or injured out of fear of discrimination.¹

Unfortunately, now, further increases to barriers to care may be put in place. In late May of this year, the Department of Health and Human Services (HHS) proposed new regulations that would reverse previous regulations granted through section 1557 of the Affordable Care Act (ACA)—the Health Care Rights Law—which affirmed the rights of gender nonbinary persons to medical care. Among the proposed changes is the elimination of protections against discrimination in health care based on gender identity.² The proposed regulation changes come on the heels of a federal court case, which seeks to declare that hospital systems may turn away patients based on gender identity.³

Unraveling rights afforded under the ACA

The Health Care Rights Law was passed under the ACA; it prohibits discrimination based on race, color, national origin, sex, age, and disability in health programs and activities receiving federal financial assistance. Multiple lower courts have supported that the rights of transgender individuals is included within these protections against discrimination on the basis of sex.⁴ These court rulings not only have ensured the ability of gender-diverse individuals to access care but also have enforced insurance coverage of therapies for gender dysphoria. It was only in 2014 that Medicaid began providing coverage for gender-affirming surgeries and eliminating language that such procedures were “experimental” or “cosmetic.” The 2016 passage of the ACA mandated that private insurance companies follow suit. Unfortunately, the recent proposed regulation changes to the Health Care Rights Law may spark a reversal from insurance companies as well. Such a setback would affect gender-diverse individuals’ hormone treatments as well as their ability to access a full spectrum of care within the health care system.

Continue to: ACOG urges nondiscriminatory practices...

ACOG urges nondiscriminatory practices

The proposed regulation changes to the Health Care Rights Law are from the Conscience and Religious Freedom Division of the HHS Office for Civil Rights, which was established in 2018 and has been advocating for the rights of health care providers to refuse to treat patients based on their own religious beliefs.⁵ We argue, however, that providing care to persons of varying backgrounds is not an assault on our individual liberties but rather a privilege as providers. As obstetrician-gynecologists, it may be easy to only consider cis-gendered women our responsibility. But our field also emphasizes individual empowerment above all else—we fight every day for our patients’ rights to contraception, fertility, pregnancy, parenthood, and sexual freedoms. Let us continue speaking up for the rights of all those who need gynecologic care, regardless of the pronouns they use.

“The American College of Obstetricians and Gynecologists urges health care providers to foster nondiscriminatory practices and policies to increase identification and to facilitate quality health care for transgender individuals, both in assisting with the transition if desired as well as providing long-term preventive health care.”⁶

We urge you to take action

• Reach out to your local representatives about protecting transgender health access
• Educate yourself on the unique needs of transgender individuals
• Read personal accounts
• Share your personal story
• Find referring providers near your practice

Vitamin E acetate is one possible culprit in the mysterious vaping-associated lung disease that has killed three patients, sickened 450, and baffled clinicians and investigators all summer.

mauro grigollo/Thinkstock

Another death may be linked to the disorder, officials said during a joint press briefing held by the Centers for Disease Control and Prevention and the Food and Drug Administration. In all, 450 potential cases have been reported and e-cigarette use confirmed in 215. Cases have occurred in 33 states and one territory. A total of 84% of the patients reported having used tetrahydrocannabinol (THC) products in e-cigarette devices.

A preliminary report on the situation by Jennifer Layden, MD, of the department of public health in Illinois and colleagues – including a preliminary case definition – was simultaneously released in the New England Journal of Medicine (2019 Sep 6. doi: 10.1056/NEJMoa1911614).

No single device or substance was common to all the cases, leading officials to issue a blanket warning against e-cigarettes, especially those containing THC.

“We believe a chemical exposure is likely related, but more information is needed to determine what substances. Some labs have identified vitamin E acetate in some samples,” said Dana Meaney-Delman, MD, MPH, incident manager, CDC 2019 Lung Injury Response. “Continued investigation is needed to identify the risk associated with a specific product or substance.”

Besides vitamin E acetate, federal labs are looking at other cannabinoids, cutting agents, diluting agents, pesticides, opioids, and toxins.

Officials also issued a general warning about the products. Youths, young people, and pregnant women should never use e-cigarettes, they cautioned, and no one should buy them from a noncertified source, a street vendor, or a social contact. Even cartridges originally obtained from a certified source should never have been altered in any way.

Dr. Layden and colleagues reported that bilateral lung infiltrates was characterized in 98% of the 53 patients hospitalized with the recently reported e-cigarette–induced lung injury. Nonspecific constitutional symptoms, including fever, chills, weight loss, and fatigue, were present in all of the patients.

Patients may show some symptoms days or even weeks before acute respiratory failure develops, and many had sought medical help before that. All presented with bilateral lung infiltrates, part of an evolving case definition. Many complained of nonspecific constitutional symptoms, including fever, chills, gastrointestinal symptoms, and weight loss. Of the patients who underwent bronchoscopy, many were diagnosed as having lipoid pneumonia, a rare condition characterized by lipid-laden macrophages.

“We don’t know the significance of the lipid-containing macrophages, and we don’t know if the lipids are endogenous or exogenous,” Dr. Meaney-Delman said.

The incidence of such cases appears to be rising rapidly, Dr. Layden noted. An epidemiologic review of cases in Illinois found that the mean monthly rate of visits related to severe respiratory illness in June-August was twice that observed during the same months last year.
 

[email protected]

SOURCE: Layden JE et al. N Engl J Med. 2019 Sep 6. doi: 1 0.1056/NEJMoa1911614.

Vitamin E acetate is one possible culprit in the mysterious vaping-associated lung disease that has killed three patients, sickened 450, and baffled clinicians and investigators all summer.

mauro grigollo/Thinkstock

Another death may be linked to the disorder, officials said during a joint press briefing held by the Centers for Disease Control and Prevention and the Food and Drug Administration. In all, 450 potential cases have been reported and e-cigarette use confirmed in 215. Cases have occurred in 33 states and one territory. A total of 84% of the patients reported having used tetrahydrocannabinol (THC) products in e-cigarette devices.

A preliminary report on the situation by Jennifer Layden, MD, of the department of public health in Illinois and colleagues – including a preliminary case definition – was simultaneously released in the New England Journal of Medicine (2019 Sep 6. doi: 10.1056/NEJMoa1911614).

No single device or substance was common to all the cases, leading officials to issue a blanket warning against e-cigarettes, especially those containing THC.

“We believe a chemical exposure is likely related, but more information is needed to determine what substances. Some labs have identified vitamin E acetate in some samples,” said Dana Meaney-Delman, MD, MPH, incident manager, CDC 2019 Lung Injury Response. “Continued investigation is needed to identify the risk associated with a specific product or substance.”

Besides vitamin E acetate, federal labs are looking at other cannabinoids, cutting agents, diluting agents, pesticides, opioids, and toxins.

Officials also issued a general warning about the products. Youths, young people, and pregnant women should never use e-cigarettes, they cautioned, and no one should buy them from a noncertified source, a street vendor, or a social contact. Even cartridges originally obtained from a certified source should never have been altered in any way.

Dr. Layden and colleagues reported that bilateral lung infiltrates was characterized in 98% of the 53 patients hospitalized with the recently reported e-cigarette–induced lung injury. Nonspecific constitutional symptoms, including fever, chills, weight loss, and fatigue, were present in all of the patients.

Patients may show some symptoms days or even weeks before acute respiratory failure develops, and many had sought medical help before that. All presented with bilateral lung infiltrates, part of an evolving case definition. Many complained of nonspecific constitutional symptoms, including fever, chills, gastrointestinal symptoms, and weight loss. Of the patients who underwent bronchoscopy, many were diagnosed as having lipoid pneumonia, a rare condition characterized by lipid-laden macrophages.

“We don’t know the significance of the lipid-containing macrophages, and we don’t know if the lipids are endogenous or exogenous,” Dr. Meaney-Delman said.

The incidence of such cases appears to be rising rapidly, Dr. Layden noted. An epidemiologic review of cases in Illinois found that the mean monthly rate of visits related to severe respiratory illness in June-August was twice that observed during the same months last year.
 

[email protected]

SOURCE: Layden JE et al. N Engl J Med. 2019 Sep 6. doi: 1 0.1056/NEJMoa1911614.

Vitamin E acetate is one possible culprit in the mysterious vaping-associated lung disease that has killed three patients, sickened 450, and baffled clinicians and investigators all summer.

mauro grigollo/Thinkstock

Another death may be linked to the disorder, officials said during a joint press briefing held by the Centers for Disease Control and Prevention and the Food and Drug Administration. In all, 450 potential cases have been reported and e-cigarette use confirmed in 215. Cases have occurred in 33 states and one territory. A total of 84% of the patients reported having used tetrahydrocannabinol (THC) products in e-cigarette devices.

A preliminary report on the situation by Jennifer Layden, MD, of the department of public health in Illinois and colleagues – including a preliminary case definition – was simultaneously released in the New England Journal of Medicine (2019 Sep 6. doi: 10.1056/NEJMoa1911614).

No single device or substance was common to all the cases, leading officials to issue a blanket warning against e-cigarettes, especially those containing THC.

“We believe a chemical exposure is likely related, but more information is needed to determine what substances. Some labs have identified vitamin E acetate in some samples,” said Dana Meaney-Delman, MD, MPH, incident manager, CDC 2019 Lung Injury Response. “Continued investigation is needed to identify the risk associated with a specific product or substance.”

Besides vitamin E acetate, federal labs are looking at other cannabinoids, cutting agents, diluting agents, pesticides, opioids, and toxins.

Officials also issued a general warning about the products. Youths, young people, and pregnant women should never use e-cigarettes, they cautioned, and no one should buy them from a noncertified source, a street vendor, or a social contact. Even cartridges originally obtained from a certified source should never have been altered in any way.

Dr. Layden and colleagues reported that bilateral lung infiltrates was characterized in 98% of the 53 patients hospitalized with the recently reported e-cigarette–induced lung injury. Nonspecific constitutional symptoms, including fever, chills, weight loss, and fatigue, were present in all of the patients.

Patients may show some symptoms days or even weeks before acute respiratory failure develops, and many had sought medical help before that. All presented with bilateral lung infiltrates, part of an evolving case definition. Many complained of nonspecific constitutional symptoms, including fever, chills, gastrointestinal symptoms, and weight loss. Of the patients who underwent bronchoscopy, many were diagnosed as having lipoid pneumonia, a rare condition characterized by lipid-laden macrophages.

“We don’t know the significance of the lipid-containing macrophages, and we don’t know if the lipids are endogenous or exogenous,” Dr. Meaney-Delman said.

The incidence of such cases appears to be rising rapidly, Dr. Layden noted. An epidemiologic review of cases in Illinois found that the mean monthly rate of visits related to severe respiratory illness in June-August was twice that observed during the same months last year.
 

[email protected]

SOURCE: Layden JE et al. N Engl J Med. 2019 Sep 6. doi: 1 0.1056/NEJMoa1911614.

Levy was chief pathologist at Veterans Health Care System of the Ozarks in Fayetteville, Arkansas. During his 12-year tenure at the US Department of Veterans Affairs (VA), he read almost 34,000 pathology slides. However, at the same time, he was working under the influence of alcohol and 2-methyl-2-butanol (2M2B)—a substance that intoxicates but cannot be detected in routine tests.

The VA fired Levy last year, and the VA Office of the Inspector General (OIG) began an investigation of his actions and of agency lapses in overseeing him. The 18-month review found that 8.9% of Levy’s diagnoses involved clinical errors—the normal misdiagnosis rate for pathologists is 0.7%. Hundreds of Levy’s misdiagnoses were not serious, but ≥ 15 may have led to deaths and harmful illness in 15 other patients. Some patients were not diagnosed when they should have been. Some were told they were sick when they were not and suffered unnecessary invasive treatment.

Levy knowingly falsified diagnoses for  3 veterans. One patient was diagnosed with diffuse large B-cell lymphoma—a type of cancer he did not have. He received the wrong treatment and died. Levy diagnosed another patient, also wrongly, with small cell carcinoma; that patient died of squamous cell carcinoma that spread. The third patient was given a benign test result for prostate cancer. Untreated, he died after the cancer spread.

One patient was given antibiotics instead of treatment for what was later diagnosed as late-stage neck and throat cancer. In an interview with the Washington Post he said, “I went from ‘Your earache isn’t anything’ to stage 4.”

How was Levy able to wreak such havoc? One reason was that despite concerns and complaints from colleagues, he looked good on paper. He falsified records to indicate that his deputy concurred with his diagnoses in mandated peer reviews. He also appeared “clean” in inspections through using 2M2B.

Levy was fired not for his work performance but for being arrested for driving while intoxicated. He had been a “star hire” with an medical degree from the University of Chicago, who had completed a pathology residency at the University of California at San Francisco and a fellowship at Duke University focusing on disease of the blood. But he also had a 1996 arrest for a driving under the influence (DUI) on his record when he joined the VA in 2005.

He was suspended again in 2017 for being under the influence but allowed to continue with nonclinical work until he was again arrested for DUI in 2018, when the police toxicology test detected 2M2B. He was finally dismissed in April 2018. Nonetheless, even after he had arrived impaired at the laboratory twice, the VA had awarded him 2 performance bonuses, based on the supposedly low clinical error rate and 42 urine and blood samples that turned up negative for alcohol and drugs.

In addition to 3 counts of involuntary manslaughter, the indictment charges that Levy devised a scheme to defraud the VA and to obtain money and property from the VA in the form of salary, benefits, and performance awards. He is charged with 12 counts of wire fraud, 12 counts of mail fraud, and 4 counts of making false statements related to 12 occasions between 2017 and 2018, when Levy was reportedly buying 2M2B over the Internet while he was contractually obligated to submit to random drug and alcohol screens.

After being fired, Levy moved to a small island in the Dutch Caribbean and found a position teaching pathology at a local medical school. At the time of his VA hiring, Levy held a medical license issued by Mississippi. His active medical licenses in California and Florida were revoked only this spring. The VA did not notify the3 states where Levy was licensed that he could no longer practice until June 2018.

The Office of Inspector General (OIG) has identified other VA physicians who continued to practice even after they were found to have compromised patient care, and the Government Accountability Office found “weak systems” for ensuring that problems are addressed in a timely fashion. A VA spokesperson, however, quoted in The Washington Post, said the Levy case was “an isolated incident,” and that the agency has “strengthened internal controls” to ensure that errors are more quickly identified and addressed. The Fayetteville Medical Center also has increased monitoring of its clinical laboratory, according to a Washington Post report. VA officials also said they have added oversight of small specialty staffs across the system to ensure “independent and objective oversight.”

The VA has contacted the families in the 30 most serious cases to advise them of their legal and treatment options, according to the Washington Post.

“The arrest of Dr. Levy was accomplished as a result of the strong leadership of the US Attorney’s Office and the extensive work of special agents of the VA OIG, supported by the medical expertise of the OIG’s health care inspection professionals,” said Michael Missal, the VA’s inspector general, in a press release issued by the US Attorney’s Office in the Western District of Arkansas. “These charges send a clear signal that anyone entrusted with the care of veterans will be held accountable for placing them at risk by working while impaired or through other misconduct.”

Levy is in jail in Fayetteville. The trial date for his case is set for October 7.

Levy was chief pathologist at Veterans Health Care System of the Ozarks in Fayetteville, Arkansas. During his 12-year tenure at the US Department of Veterans Affairs (VA), he read almost 34,000 pathology slides. However, at the same time, he was working under the influence of alcohol and 2-methyl-2-butanol (2M2B)—a substance that intoxicates but cannot be detected in routine tests.

The VA fired Levy last year, and the VA Office of the Inspector General (OIG) began an investigation of his actions and of agency lapses in overseeing him. The 18-month review found that 8.9% of Levy’s diagnoses involved clinical errors—the normal misdiagnosis rate for pathologists is 0.7%. Hundreds of Levy’s misdiagnoses were not serious, but ≥ 15 may have led to deaths and harmful illness in 15 other patients. Some patients were not diagnosed when they should have been. Some were told they were sick when they were not and suffered unnecessary invasive treatment.

Levy knowingly falsified diagnoses for  3 veterans. One patient was diagnosed with diffuse large B-cell lymphoma—a type of cancer he did not have. He received the wrong treatment and died. Levy diagnosed another patient, also wrongly, with small cell carcinoma; that patient died of squamous cell carcinoma that spread. The third patient was given a benign test result for prostate cancer. Untreated, he died after the cancer spread.

One patient was given antibiotics instead of treatment for what was later diagnosed as late-stage neck and throat cancer. In an interview with the Washington Post he said, “I went from ‘Your earache isn’t anything’ to stage 4.”

How was Levy able to wreak such havoc? One reason was that despite concerns and complaints from colleagues, he looked good on paper. He falsified records to indicate that his deputy concurred with his diagnoses in mandated peer reviews. He also appeared “clean” in inspections through using 2M2B.

Levy was fired not for his work performance but for being arrested for driving while intoxicated. He had been a “star hire” with an medical degree from the University of Chicago, who had completed a pathology residency at the University of California at San Francisco and a fellowship at Duke University focusing on disease of the blood. But he also had a 1996 arrest for a driving under the influence (DUI) on his record when he joined the VA in 2005.

He was suspended again in 2017 for being under the influence but allowed to continue with nonclinical work until he was again arrested for DUI in 2018, when the police toxicology test detected 2M2B. He was finally dismissed in April 2018. Nonetheless, even after he had arrived impaired at the laboratory twice, the VA had awarded him 2 performance bonuses, based on the supposedly low clinical error rate and 42 urine and blood samples that turned up negative for alcohol and drugs.

In addition to 3 counts of involuntary manslaughter, the indictment charges that Levy devised a scheme to defraud the VA and to obtain money and property from the VA in the form of salary, benefits, and performance awards. He is charged with 12 counts of wire fraud, 12 counts of mail fraud, and 4 counts of making false statements related to 12 occasions between 2017 and 2018, when Levy was reportedly buying 2M2B over the Internet while he was contractually obligated to submit to random drug and alcohol screens.

After being fired, Levy moved to a small island in the Dutch Caribbean and found a position teaching pathology at a local medical school. At the time of his VA hiring, Levy held a medical license issued by Mississippi. His active medical licenses in California and Florida were revoked only this spring. The VA did not notify the3 states where Levy was licensed that he could no longer practice until June 2018.

The Office of Inspector General (OIG) has identified other VA physicians who continued to practice even after they were found to have compromised patient care, and the Government Accountability Office found “weak systems” for ensuring that problems are addressed in a timely fashion. A VA spokesperson, however, quoted in The Washington Post, said the Levy case was “an isolated incident,” and that the agency has “strengthened internal controls” to ensure that errors are more quickly identified and addressed. The Fayetteville Medical Center also has increased monitoring of its clinical laboratory, according to a Washington Post report. VA officials also said they have added oversight of small specialty staffs across the system to ensure “independent and objective oversight.”

The VA has contacted the families in the 30 most serious cases to advise them of their legal and treatment options, according to the Washington Post.

“The arrest of Dr. Levy was accomplished as a result of the strong leadership of the US Attorney’s Office and the extensive work of special agents of the VA OIG, supported by the medical expertise of the OIG’s health care inspection professionals,” said Michael Missal, the VA’s inspector general, in a press release issued by the US Attorney’s Office in the Western District of Arkansas. “These charges send a clear signal that anyone entrusted with the care of veterans will be held accountable for placing them at risk by working while impaired or through other misconduct.”

Levy is in jail in Fayetteville. The trial date for his case is set for October 7.

Levy was chief pathologist at Veterans Health Care System of the Ozarks in Fayetteville, Arkansas. During his 12-year tenure at the US Department of Veterans Affairs (VA), he read almost 34,000 pathology slides. However, at the same time, he was working under the influence of alcohol and 2-methyl-2-butanol (2M2B)—a substance that intoxicates but cannot be detected in routine tests.

The VA fired Levy last year, and the VA Office of the Inspector General (OIG) began an investigation of his actions and of agency lapses in overseeing him. The 18-month review found that 8.9% of Levy’s diagnoses involved clinical errors—the normal misdiagnosis rate for pathologists is 0.7%. Hundreds of Levy’s misdiagnoses were not serious, but ≥ 15 may have led to deaths and harmful illness in 15 other patients. Some patients were not diagnosed when they should have been. Some were told they were sick when they were not and suffered unnecessary invasive treatment.

Levy knowingly falsified diagnoses for  3 veterans. One patient was diagnosed with diffuse large B-cell lymphoma—a type of cancer he did not have. He received the wrong treatment and died. Levy diagnosed another patient, also wrongly, with small cell carcinoma; that patient died of squamous cell carcinoma that spread. The third patient was given a benign test result for prostate cancer. Untreated, he died after the cancer spread.

One patient was given antibiotics instead of treatment for what was later diagnosed as late-stage neck and throat cancer. In an interview with the Washington Post he said, “I went from ‘Your earache isn’t anything’ to stage 4.”

How was Levy able to wreak such havoc? One reason was that despite concerns and complaints from colleagues, he looked good on paper. He falsified records to indicate that his deputy concurred with his diagnoses in mandated peer reviews. He also appeared “clean” in inspections through using 2M2B.

Levy was fired not for his work performance but for being arrested for driving while intoxicated. He had been a “star hire” with an medical degree from the University of Chicago, who had completed a pathology residency at the University of California at San Francisco and a fellowship at Duke University focusing on disease of the blood. But he also had a 1996 arrest for a driving under the influence (DUI) on his record when he joined the VA in 2005.

He was suspended again in 2017 for being under the influence but allowed to continue with nonclinical work until he was again arrested for DUI in 2018, when the police toxicology test detected 2M2B. He was finally dismissed in April 2018. Nonetheless, even after he had arrived impaired at the laboratory twice, the VA had awarded him 2 performance bonuses, based on the supposedly low clinical error rate and 42 urine and blood samples that turned up negative for alcohol and drugs.

In addition to 3 counts of involuntary manslaughter, the indictment charges that Levy devised a scheme to defraud the VA and to obtain money and property from the VA in the form of salary, benefits, and performance awards. He is charged with 12 counts of wire fraud, 12 counts of mail fraud, and 4 counts of making false statements related to 12 occasions between 2017 and 2018, when Levy was reportedly buying 2M2B over the Internet while he was contractually obligated to submit to random drug and alcohol screens.

After being fired, Levy moved to a small island in the Dutch Caribbean and found a position teaching pathology at a local medical school. At the time of his VA hiring, Levy held a medical license issued by Mississippi. His active medical licenses in California and Florida were revoked only this spring. The VA did not notify the3 states where Levy was licensed that he could no longer practice until June 2018.

The Office of Inspector General (OIG) has identified other VA physicians who continued to practice even after they were found to have compromised patient care, and the Government Accountability Office found “weak systems” for ensuring that problems are addressed in a timely fashion. A VA spokesperson, however, quoted in The Washington Post, said the Levy case was “an isolated incident,” and that the agency has “strengthened internal controls” to ensure that errors are more quickly identified and addressed. The Fayetteville Medical Center also has increased monitoring of its clinical laboratory, according to a Washington Post report. VA officials also said they have added oversight of small specialty staffs across the system to ensure “independent and objective oversight.”

The VA has contacted the families in the 30 most serious cases to advise them of their legal and treatment options, according to the Washington Post.

“The arrest of Dr. Levy was accomplished as a result of the strong leadership of the US Attorney’s Office and the extensive work of special agents of the VA OIG, supported by the medical expertise of the OIG’s health care inspection professionals,” said Michael Missal, the VA’s inspector general, in a press release issued by the US Attorney’s Office in the Western District of Arkansas. “These charges send a clear signal that anyone entrusted with the care of veterans will be held accountable for placing them at risk by working while impaired or through other misconduct.”

Levy is in jail in Fayetteville. The trial date for his case is set for October 7.

The Federal Bureau of Labor Statistics projects 37% job growth for physician assistants (PAs) from 2016 to 2026, much greater than the average for all other occupations as well as for other medical professions.¹ This growth has been accompanied by increased enrollment in medical (doctor of medicine [MD], doctor of osteopathic medicine) and nurse practitioner (NP) schools.² Clinical teaching sites serve a crucial function in the training of all clinical disciplines. These sites provide hands-on and experiential learning in medical settings, necessary components for learners practicing to become clinicians. Significant PA program expansion has led to increased demand for clinical training, creating competition for sites and a shortage of willing and well-trained preceptors.³

This challenge has been recognized by PA program directors. In the Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey, PA program directors expressed concern about the adequacy of clinical opportunities for students, increased difficulty developing new core sites, and preserving existing core sites. In addition, they noted that a shortage of clinical sites was one of the greatest barriers to the PA programs’ sustained growth and success.⁴

Program directors also indicated difficulty securing clinical training sites in internal medicine (IM) and high rates of attrition of medicine clinical preceptors for their students.⁵ The reasons are multifold: increasing clinical demands, time, teaching competence, lack of experience, academic affiliation, lack of reimbursement, or compensation. Moreover, there is a declining number of PAs who work in primary care compared with specialty and subspecialty care, limiting the availability of clinical training preceptors in medicine and primary care.^6-8 According to the American Academy of PAs (AAPA) census and salary survey data, the percentage of PAs working in the primary care specialties (ie, family medicine, IM, and general pediatrics) has decreased from > 47% in 1995 to 24% in 2017.⁹ As such, there is a need to broaden the educational landscape to provide more high-quality training sites in IM.

The postacute health care setting may address this training need. It offers a unique clinical opportunity to expose learners to a broad range of disease complexity and clinical acuity, as the percentage of patients discharged from hospitals to postacute care (PAC) has increased and care shifts from the hospital to the PAC setting.^10,11 The longer PAC length of stay also enables learners to follow patients longitudinally over several weeks and experience interprofessional team-based care. In addition, the PAC setting offers learners the ability to acquire the necessary skills for smooth and effective transitions of care. This setting has been extensively used for trainees of nursing, pharmacy, physical therapy (PT) and occupational therapy (OT), speech-language pathology, psychology, and social work (SW), but few programs have used the PAC setting as clerkship sites for IM rotations for PA students. To address this need for IM sites, the VA Boston Healthcare System (VABHS), in conjunction with the Boston University School of Medicine Physician Assistant Program, developed a novel medicine clinical clerkship site for physician assistants in the PAC unit of the community living center (CLC) at VABHS. This report describes the program structure, curriculum, and participant evaluation results.

Clinical Clerkship Program

VABHS CLC is a 110-bed facility comprising 3 units: a 65-bed PAC unit, a 15-bed closed hospice/palliative care unit, and a 30-bed long-term care unit. The service is staffed continuously with physicians, PAs, and NPs. A majority of patients are admitted from the acute care hospital of VABHS (West Roxbury campus) and other regional VA facilities. The CLC offers dynamic services, including phlebotomy, general radiology, IV diuretics and antibiotics, wound care, and subacute PT, OT, and speech-language pathology rehabilitation. The CLC serves as a venue for transitioning patients from acute inpatient care to home. The patient population is often elderly, with multiple active comorbidities and variable medical literacy, adherence, and follow-up.

The CLC provides a diverse interprofessional learning environment, offering core IM rotations for first-year psychiatry residents, oral and maxillofacial surgery residents, and PA students. The CLC also has expanded as a clinical site both for transitions-in-care IM resident curricula and electives as well as a geriatrics fellowship. In addition, the site offers rotations for NPs, nursing, pharmacy, physical and occupational therapies, speech-language pathology, psychology, and SW.

The Boston University School of Medicine Physician Assistant Program was founded in 2015 as a master’s degree program completed over 28 months. The first 12 months are didactic, and the following 16 months are clinical training with 14 months of rotations (2 IM, family medicine, pediatrics, emergency medicine, general surgery, obstetrics and gynecology, psychiatry, neurology, and 5 elective rotations), and 2 months for a thesis. The program has about 30 students per year and 4 clerkship sites for IM.

Program Description

The VABHS medicine clerkship hosts 1 to 2 PA students for 4-week blocks in the PAC unit of the CLC. Each student rotates on both PA and MD teams. Students follow 3 to 4 patients and participate fully in their care from admission to discharge; they prepare daily presentations and participate in medical management, family meetings, chart documentation, and care coordination with the interprofessional team. Students are provided a physical examination checklist and feedback form, and they are expected to track findings and record feedback and goals with their supervising preceptor weekly. They also make formal case presentations and participate in monthly medicine didactic rounds available to all VABHS IM students and trainees via videoconference.

In addition, beginning in July 2017, all PA students in the CLC began to participate in a 4-week Interprofessional Curriculum in Transitional Care. The curriculum includes 14 didactic lectures taught by 16 interprofessional faculty, including medicine, geriatric, and palliative care physicians; PAs; social workers; physical and occupational therapists; pharmacists; and a geriatric psychologist. The didactics include topics on the interprofessional team, the care continuum, teams and teamwork, interdisciplinary coordination of care, components of effective transitions in care, medication reconciliation, approaching difficult conversations, advance care planning, and quality improvement. The goal of the curriculum is to provide learners the knowledge, skills, and dispositions necessary for high-quality transitional care and interprofessional practice as well as specific training for effective and safe transfers of care between clinical settings. Although PA students are the main participants in this curriculum, all other learners in the PAC unit are also invited to attend the lectures.

The unique attributes of this training site include direct interaction with supervising PAs and physicians, rather than experiencing the traditional teaching hierarchy (with interns, residents, fellows); observation of the natural progression of disease of both acute care and primary care issues due to the longer length of stay (2 to 6 weeks, where the typical student will see the same patient 7 to 10 times during their rotation); exposure to a host of medically complex patients offering a multitude of clinical scenarios and abnormal physical exam findings; exposure to a hospice/palliative care ward and end-of-life care; and interaction within an interprofessional training environment of nursing, pharmacy, PT, OT, speech-language pathology, psychology, and SW trainees.

Program Evaluation

At the end of rotations continuously through the year, PA students electronically complete a site evaluation from the Boston University School of Medicine Physician Assistant Program. The evaluation consists of 14 questions: 6 about site quality and 8 about instruction quality. The questions are answered on a 5-point Likert scale. Also included are 2 open-ended response questions that ask what they liked about the rotation and what they felt could be improved. Results are anonymous, de-identified and blinded both to the program as well as the clerkship site. Results are aggregated and provided to program sites annually. Responses are converted to a dichotomous variable, where any good or excellent response (4 or 5) is considered positive and any neutral or below (3, 2, 1) is considered a nonpositive response.

Results

The clerkship site has been operational since June 22, 2015. There have been 59 students who participated in the rotation. A different scale in these evaluations was used between June 22, 2015, and September 13, 2015. Therefore, 7 responses were excluded from the analysis, leaving 52 usable evaluations. The responses were analyzed both in total (for the CLC as well as other IM rotation sites) and by individual clerkship year to look for any trends over time: September 14, 2015, through April 24, 2016; April 25, 2016, through April 28, 2017; and May 1, 2017, through March 1, 2018 (Table).

Site evaluations showed high satisfaction regarding the quality of the physical environment as well as the learning environment. Students endorsed the PAC unit having resources and physical space for them, such as a desk and computer, opportunity for participation in patient care, and parking (100%; n = 52). Site evaluations revealed high satisfaction with the quality of teaching and faculty encouragement and support of their learning (100%; n = 52). The evaluations revealed that bedside teaching was strong (94%; n = 49). The students reported high satisfaction with the volume of patients provided (92%; n = 48) as well as the diversity of diagnoses (92%; n = 48).

There were fewer positive responses in the first 2 years of the rotation with regard to formal lectures (50% and 67%; 7/14 and 16/24, respectively). In the third year of the rotation, students had a much higher satisfaction rate (93%; 13/14). This increased satisfaction was associated with the development and incorporation of the Interprofessional Curriculum in Transitional Care in 2017.

Discussion

Access to high-quality PA student clerkship sites has become a pressing issue in recent years because of increased competition for sites and a shortage of willing and well-trained preceptors. There has been marked growth in schools and enrollment across all medical professions. The Accreditation Review Commission on Education for the PA (ARC-PA) reported that the total number of accredited entry-level PA programs in 2018 was 246, with 58 new accredited programs projected by 2022.¹² The Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey reported a 66% increase in first-year enrollment in PA programs from 2002 to 2012.⁵ Programs must implement alternative strategies to attract clinical sites (eg, academic appointments, increased clinical resources to training sites) or face continued challenges with recruiting training sites for their students. Postacute care may be a natural extension to expand the footprint for clinical sites for these programs, augmenting acute inpatient and outpatient rotations. This implementation would increase the pool of clinical training sites and preceptors.

The experience with this novel training site, based on PA student feedback and evaluations, has been positive, and the postacute setting can provide students with high-quality IM clinical experiences. Students report adequate patient volume and diversity. In addition, evaluations are comparable with that of other IM site rotations the students experience. Qualitative feedback has emphasized the value of following patients over longer periods; eg, weeks vs days (as in acute care) enabling students to build relationships with patients as well as observe a richer clinical spectrum of disease over a less compressed period. “Patients have complex issues, so from a medical standpoint it challenges you to think of new ways to manage their care,” commented a representative student. “It is really beneficial that you can follow them over time.”

Furthermore, in response to student feedback on didactics, an interprofessional curriculum was developed to add formal structure as well as to create a curriculum in care transitions. This curriculum provided a unique opportunity for PA students to receive formal instruction on areas of particular relevance for transitional care (eg, care continuum, end of life issues, and care transitions). The curriculum also allows the interprofessional faculty a unique and enjoyable opportunity for interprofessional collaboration.

The 1 month PAC rotation is augmented with inpatient IM and outpatient family medicine rotations, consequently giving exposure to the full continuum of care. The PAC setting provides learners multifaceted benefits: the opportunity to strengthen and develop the knowledge, attitudes, and skills necessary for IM; increased understanding of other professions by observing and interacting as a team caring for a patient over a longer period as opposed to the acute care setting; the ability to perform effective, efficient, and safe transfer between clinical settings; and broad exposure to transitional care. As a result, the PAC rotation enhances but does not replace the necessary and essential rotations of inpatient and outpatient medicine.

Moreover, this rotation provides unique and core IM training for PA students. Our site focuses on interprofessional collaboration, emphasizing the importance of team-based care, an essential concept in modern day medicine. Formal exposure to other care specialties, such as PT and OT, SW, and mental health, is essential for students to appreciate clinical medicine and a patient’s physical and mental experience over the course of a disease and clinical state. In addition, the physical exam checklist ensures that students are exposed to the full spectrum of IM examination findings during their rotation. Finally, weekly feedback forms require students to ask and receive concrete feedback from their supervising providers.

Limitations

The generalizability of this model requires careful consideration. VABHS is a tertiary care integrated health care system, enabling students to learn from patients moving through multiple care transitions in a single health care system. In addition, other settings may not have the staffing or clinical volume to sustain such a model. All PAC clinical faculty teach voluntarily, and local leadership has set expectations for all clinicians to participate in teaching of trainees and PA students. Evaluations also note less diversity in the patient population, a challenge that some VA facilities face. This issue could be addressed by ensuring that students also have IM rotations at other inpatient medical facilities. A more balanced experience, where students reap the positive benefits of PAC but do not lose exposure to a diverse patient pool, could result. Furthermore, some of the perceived positive impacts also may be related to professional and personal attributes of the teaching clinicians rather than to the PAC setting.

Conclusion

PAC settings can be effective training sites for medicine clerkships for PA students and can provide high-quality training in IM as PA programs continue to expand. This setting offers students exposure to interprofessional, team-based care and the opportunity to care for patients with a broad range of disease complexity. Learning is further enhanced by the ability to follow patients longitudinally over their disease course as well as to work directly with teaching faculty and other interprofessional health care professionals. Evaluations of this novel clerkship experience have shown high levels of student satisfaction in knowledge growth, clinical skills, bedside teaching, and mentorship.

Acknowledgments
We thank Juman Hijab for her critical role in establishing and maintaining the clerkship. We thank Steven Simon, Matt Russell, and Thomas Parrino for their leadership and guidance in establishing and maintaining the clerkship. We thank the Boston University School of Medicine Physician Assistant Program Director Mary Warner for her support and guidance in creating and supporting the clerkship. In addition, we thank the interprofessional education faculty for their dedicated involvement in teaching, including Stephanie Saunders, Lindsay Lefers, Jessica Rawlins, Lindsay Brennan, Angela Viani, Eric Charette, Nicole O’Neil, Susan Nathan, Jordana Meyerson, Shivani Jindal, Wei Shen, Amy Hanson, Gilda Cain, and Kate Hinrichs.

The Federal Bureau of Labor Statistics projects 37% job growth for physician assistants (PAs) from 2016 to 2026, much greater than the average for all other occupations as well as for other medical professions.¹ This growth has been accompanied by increased enrollment in medical (doctor of medicine [MD], doctor of osteopathic medicine) and nurse practitioner (NP) schools.² Clinical teaching sites serve a crucial function in the training of all clinical disciplines. These sites provide hands-on and experiential learning in medical settings, necessary components for learners practicing to become clinicians. Significant PA program expansion has led to increased demand for clinical training, creating competition for sites and a shortage of willing and well-trained preceptors.³

This challenge has been recognized by PA program directors. In the Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey, PA program directors expressed concern about the adequacy of clinical opportunities for students, increased difficulty developing new core sites, and preserving existing core sites. In addition, they noted that a shortage of clinical sites was one of the greatest barriers to the PA programs’ sustained growth and success.⁴

Program directors also indicated difficulty securing clinical training sites in internal medicine (IM) and high rates of attrition of medicine clinical preceptors for their students.⁵ The reasons are multifold: increasing clinical demands, time, teaching competence, lack of experience, academic affiliation, lack of reimbursement, or compensation. Moreover, there is a declining number of PAs who work in primary care compared with specialty and subspecialty care, limiting the availability of clinical training preceptors in medicine and primary care.^6-8 According to the American Academy of PAs (AAPA) census and salary survey data, the percentage of PAs working in the primary care specialties (ie, family medicine, IM, and general pediatrics) has decreased from > 47% in 1995 to 24% in 2017.⁹ As such, there is a need to broaden the educational landscape to provide more high-quality training sites in IM.

The postacute health care setting may address this training need. It offers a unique clinical opportunity to expose learners to a broad range of disease complexity and clinical acuity, as the percentage of patients discharged from hospitals to postacute care (PAC) has increased and care shifts from the hospital to the PAC setting.^10,11 The longer PAC length of stay also enables learners to follow patients longitudinally over several weeks and experience interprofessional team-based care. In addition, the PAC setting offers learners the ability to acquire the necessary skills for smooth and effective transitions of care. This setting has been extensively used for trainees of nursing, pharmacy, physical therapy (PT) and occupational therapy (OT), speech-language pathology, psychology, and social work (SW), but few programs have used the PAC setting as clerkship sites for IM rotations for PA students. To address this need for IM sites, the VA Boston Healthcare System (VABHS), in conjunction with the Boston University School of Medicine Physician Assistant Program, developed a novel medicine clinical clerkship site for physician assistants in the PAC unit of the community living center (CLC) at VABHS. This report describes the program structure, curriculum, and participant evaluation results.

Clinical Clerkship Program

VABHS CLC is a 110-bed facility comprising 3 units: a 65-bed PAC unit, a 15-bed closed hospice/palliative care unit, and a 30-bed long-term care unit. The service is staffed continuously with physicians, PAs, and NPs. A majority of patients are admitted from the acute care hospital of VABHS (West Roxbury campus) and other regional VA facilities. The CLC offers dynamic services, including phlebotomy, general radiology, IV diuretics and antibiotics, wound care, and subacute PT, OT, and speech-language pathology rehabilitation. The CLC serves as a venue for transitioning patients from acute inpatient care to home. The patient population is often elderly, with multiple active comorbidities and variable medical literacy, adherence, and follow-up.

The CLC provides a diverse interprofessional learning environment, offering core IM rotations for first-year psychiatry residents, oral and maxillofacial surgery residents, and PA students. The CLC also has expanded as a clinical site both for transitions-in-care IM resident curricula and electives as well as a geriatrics fellowship. In addition, the site offers rotations for NPs, nursing, pharmacy, physical and occupational therapies, speech-language pathology, psychology, and SW.

The Boston University School of Medicine Physician Assistant Program was founded in 2015 as a master’s degree program completed over 28 months. The first 12 months are didactic, and the following 16 months are clinical training with 14 months of rotations (2 IM, family medicine, pediatrics, emergency medicine, general surgery, obstetrics and gynecology, psychiatry, neurology, and 5 elective rotations), and 2 months for a thesis. The program has about 30 students per year and 4 clerkship sites for IM.

Program Description

The VABHS medicine clerkship hosts 1 to 2 PA students for 4-week blocks in the PAC unit of the CLC. Each student rotates on both PA and MD teams. Students follow 3 to 4 patients and participate fully in their care from admission to discharge; they prepare daily presentations and participate in medical management, family meetings, chart documentation, and care coordination with the interprofessional team. Students are provided a physical examination checklist and feedback form, and they are expected to track findings and record feedback and goals with their supervising preceptor weekly. They also make formal case presentations and participate in monthly medicine didactic rounds available to all VABHS IM students and trainees via videoconference.

In addition, beginning in July 2017, all PA students in the CLC began to participate in a 4-week Interprofessional Curriculum in Transitional Care. The curriculum includes 14 didactic lectures taught by 16 interprofessional faculty, including medicine, geriatric, and palliative care physicians; PAs; social workers; physical and occupational therapists; pharmacists; and a geriatric psychologist. The didactics include topics on the interprofessional team, the care continuum, teams and teamwork, interdisciplinary coordination of care, components of effective transitions in care, medication reconciliation, approaching difficult conversations, advance care planning, and quality improvement. The goal of the curriculum is to provide learners the knowledge, skills, and dispositions necessary for high-quality transitional care and interprofessional practice as well as specific training for effective and safe transfers of care between clinical settings. Although PA students are the main participants in this curriculum, all other learners in the PAC unit are also invited to attend the lectures.

The unique attributes of this training site include direct interaction with supervising PAs and physicians, rather than experiencing the traditional teaching hierarchy (with interns, residents, fellows); observation of the natural progression of disease of both acute care and primary care issues due to the longer length of stay (2 to 6 weeks, where the typical student will see the same patient 7 to 10 times during their rotation); exposure to a host of medically complex patients offering a multitude of clinical scenarios and abnormal physical exam findings; exposure to a hospice/palliative care ward and end-of-life care; and interaction within an interprofessional training environment of nursing, pharmacy, PT, OT, speech-language pathology, psychology, and SW trainees.

Program Evaluation

At the end of rotations continuously through the year, PA students electronically complete a site evaluation from the Boston University School of Medicine Physician Assistant Program. The evaluation consists of 14 questions: 6 about site quality and 8 about instruction quality. The questions are answered on a 5-point Likert scale. Also included are 2 open-ended response questions that ask what they liked about the rotation and what they felt could be improved. Results are anonymous, de-identified and blinded both to the program as well as the clerkship site. Results are aggregated and provided to program sites annually. Responses are converted to a dichotomous variable, where any good or excellent response (4 or 5) is considered positive and any neutral or below (3, 2, 1) is considered a nonpositive response.

Results

The clerkship site has been operational since June 22, 2015. There have been 59 students who participated in the rotation. A different scale in these evaluations was used between June 22, 2015, and September 13, 2015. Therefore, 7 responses were excluded from the analysis, leaving 52 usable evaluations. The responses were analyzed both in total (for the CLC as well as other IM rotation sites) and by individual clerkship year to look for any trends over time: September 14, 2015, through April 24, 2016; April 25, 2016, through April 28, 2017; and May 1, 2017, through March 1, 2018 (Table).

Site evaluations showed high satisfaction regarding the quality of the physical environment as well as the learning environment. Students endorsed the PAC unit having resources and physical space for them, such as a desk and computer, opportunity for participation in patient care, and parking (100%; n = 52). Site evaluations revealed high satisfaction with the quality of teaching and faculty encouragement and support of their learning (100%; n = 52). The evaluations revealed that bedside teaching was strong (94%; n = 49). The students reported high satisfaction with the volume of patients provided (92%; n = 48) as well as the diversity of diagnoses (92%; n = 48).

There were fewer positive responses in the first 2 years of the rotation with regard to formal lectures (50% and 67%; 7/14 and 16/24, respectively). In the third year of the rotation, students had a much higher satisfaction rate (93%; 13/14). This increased satisfaction was associated with the development and incorporation of the Interprofessional Curriculum in Transitional Care in 2017.

Discussion

Access to high-quality PA student clerkship sites has become a pressing issue in recent years because of increased competition for sites and a shortage of willing and well-trained preceptors. There has been marked growth in schools and enrollment across all medical professions. The Accreditation Review Commission on Education for the PA (ARC-PA) reported that the total number of accredited entry-level PA programs in 2018 was 246, with 58 new accredited programs projected by 2022.¹² The Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey reported a 66% increase in first-year enrollment in PA programs from 2002 to 2012.⁵ Programs must implement alternative strategies to attract clinical sites (eg, academic appointments, increased clinical resources to training sites) or face continued challenges with recruiting training sites for their students. Postacute care may be a natural extension to expand the footprint for clinical sites for these programs, augmenting acute inpatient and outpatient rotations. This implementation would increase the pool of clinical training sites and preceptors.

The experience with this novel training site, based on PA student feedback and evaluations, has been positive, and the postacute setting can provide students with high-quality IM clinical experiences. Students report adequate patient volume and diversity. In addition, evaluations are comparable with that of other IM site rotations the students experience. Qualitative feedback has emphasized the value of following patients over longer periods; eg, weeks vs days (as in acute care) enabling students to build relationships with patients as well as observe a richer clinical spectrum of disease over a less compressed period. “Patients have complex issues, so from a medical standpoint it challenges you to think of new ways to manage their care,” commented a representative student. “It is really beneficial that you can follow them over time.”

Furthermore, in response to student feedback on didactics, an interprofessional curriculum was developed to add formal structure as well as to create a curriculum in care transitions. This curriculum provided a unique opportunity for PA students to receive formal instruction on areas of particular relevance for transitional care (eg, care continuum, end of life issues, and care transitions). The curriculum also allows the interprofessional faculty a unique and enjoyable opportunity for interprofessional collaboration.

The 1 month PAC rotation is augmented with inpatient IM and outpatient family medicine rotations, consequently giving exposure to the full continuum of care. The PAC setting provides learners multifaceted benefits: the opportunity to strengthen and develop the knowledge, attitudes, and skills necessary for IM; increased understanding of other professions by observing and interacting as a team caring for a patient over a longer period as opposed to the acute care setting; the ability to perform effective, efficient, and safe transfer between clinical settings; and broad exposure to transitional care. As a result, the PAC rotation enhances but does not replace the necessary and essential rotations of inpatient and outpatient medicine.

Moreover, this rotation provides unique and core IM training for PA students. Our site focuses on interprofessional collaboration, emphasizing the importance of team-based care, an essential concept in modern day medicine. Formal exposure to other care specialties, such as PT and OT, SW, and mental health, is essential for students to appreciate clinical medicine and a patient’s physical and mental experience over the course of a disease and clinical state. In addition, the physical exam checklist ensures that students are exposed to the full spectrum of IM examination findings during their rotation. Finally, weekly feedback forms require students to ask and receive concrete feedback from their supervising providers.

Limitations

The generalizability of this model requires careful consideration. VABHS is a tertiary care integrated health care system, enabling students to learn from patients moving through multiple care transitions in a single health care system. In addition, other settings may not have the staffing or clinical volume to sustain such a model. All PAC clinical faculty teach voluntarily, and local leadership has set expectations for all clinicians to participate in teaching of trainees and PA students. Evaluations also note less diversity in the patient population, a challenge that some VA facilities face. This issue could be addressed by ensuring that students also have IM rotations at other inpatient medical facilities. A more balanced experience, where students reap the positive benefits of PAC but do not lose exposure to a diverse patient pool, could result. Furthermore, some of the perceived positive impacts also may be related to professional and personal attributes of the teaching clinicians rather than to the PAC setting.

Conclusion

PAC settings can be effective training sites for medicine clerkships for PA students and can provide high-quality training in IM as PA programs continue to expand. This setting offers students exposure to interprofessional, team-based care and the opportunity to care for patients with a broad range of disease complexity. Learning is further enhanced by the ability to follow patients longitudinally over their disease course as well as to work directly with teaching faculty and other interprofessional health care professionals. Evaluations of this novel clerkship experience have shown high levels of student satisfaction in knowledge growth, clinical skills, bedside teaching, and mentorship.

Acknowledgments
We thank Juman Hijab for her critical role in establishing and maintaining the clerkship. We thank Steven Simon, Matt Russell, and Thomas Parrino for their leadership and guidance in establishing and maintaining the clerkship. We thank the Boston University School of Medicine Physician Assistant Program Director Mary Warner for her support and guidance in creating and supporting the clerkship. In addition, we thank the interprofessional education faculty for their dedicated involvement in teaching, including Stephanie Saunders, Lindsay Lefers, Jessica Rawlins, Lindsay Brennan, Angela Viani, Eric Charette, Nicole O’Neil, Susan Nathan, Jordana Meyerson, Shivani Jindal, Wei Shen, Amy Hanson, Gilda Cain, and Kate Hinrichs.

The Federal Bureau of Labor Statistics projects 37% job growth for physician assistants (PAs) from 2016 to 2026, much greater than the average for all other occupations as well as for other medical professions.¹ This growth has been accompanied by increased enrollment in medical (doctor of medicine [MD], doctor of osteopathic medicine) and nurse practitioner (NP) schools.² Clinical teaching sites serve a crucial function in the training of all clinical disciplines. These sites provide hands-on and experiential learning in medical settings, necessary components for learners practicing to become clinicians. Significant PA program expansion has led to increased demand for clinical training, creating competition for sites and a shortage of willing and well-trained preceptors.³

This challenge has been recognized by PA program directors. In the Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey, PA program directors expressed concern about the adequacy of clinical opportunities for students, increased difficulty developing new core sites, and preserving existing core sites. In addition, they noted that a shortage of clinical sites was one of the greatest barriers to the PA programs’ sustained growth and success.⁴

Program directors also indicated difficulty securing clinical training sites in internal medicine (IM) and high rates of attrition of medicine clinical preceptors for their students.⁵ The reasons are multifold: increasing clinical demands, time, teaching competence, lack of experience, academic affiliation, lack of reimbursement, or compensation. Moreover, there is a declining number of PAs who work in primary care compared with specialty and subspecialty care, limiting the availability of clinical training preceptors in medicine and primary care.^6-8 According to the American Academy of PAs (AAPA) census and salary survey data, the percentage of PAs working in the primary care specialties (ie, family medicine, IM, and general pediatrics) has decreased from > 47% in 1995 to 24% in 2017.⁹ As such, there is a need to broaden the educational landscape to provide more high-quality training sites in IM.

The postacute health care setting may address this training need. It offers a unique clinical opportunity to expose learners to a broad range of disease complexity and clinical acuity, as the percentage of patients discharged from hospitals to postacute care (PAC) has increased and care shifts from the hospital to the PAC setting.^10,11 The longer PAC length of stay also enables learners to follow patients longitudinally over several weeks and experience interprofessional team-based care. In addition, the PAC setting offers learners the ability to acquire the necessary skills for smooth and effective transitions of care. This setting has been extensively used for trainees of nursing, pharmacy, physical therapy (PT) and occupational therapy (OT), speech-language pathology, psychology, and social work (SW), but few programs have used the PAC setting as clerkship sites for IM rotations for PA students. To address this need for IM sites, the VA Boston Healthcare System (VABHS), in conjunction with the Boston University School of Medicine Physician Assistant Program, developed a novel medicine clinical clerkship site for physician assistants in the PAC unit of the community living center (CLC) at VABHS. This report describes the program structure, curriculum, and participant evaluation results.

Clinical Clerkship Program

VABHS CLC is a 110-bed facility comprising 3 units: a 65-bed PAC unit, a 15-bed closed hospice/palliative care unit, and a 30-bed long-term care unit. The service is staffed continuously with physicians, PAs, and NPs. A majority of patients are admitted from the acute care hospital of VABHS (West Roxbury campus) and other regional VA facilities. The CLC offers dynamic services, including phlebotomy, general radiology, IV diuretics and antibiotics, wound care, and subacute PT, OT, and speech-language pathology rehabilitation. The CLC serves as a venue for transitioning patients from acute inpatient care to home. The patient population is often elderly, with multiple active comorbidities and variable medical literacy, adherence, and follow-up.

The CLC provides a diverse interprofessional learning environment, offering core IM rotations for first-year psychiatry residents, oral and maxillofacial surgery residents, and PA students. The CLC also has expanded as a clinical site both for transitions-in-care IM resident curricula and electives as well as a geriatrics fellowship. In addition, the site offers rotations for NPs, nursing, pharmacy, physical and occupational therapies, speech-language pathology, psychology, and SW.

The Boston University School of Medicine Physician Assistant Program was founded in 2015 as a master’s degree program completed over 28 months. The first 12 months are didactic, and the following 16 months are clinical training with 14 months of rotations (2 IM, family medicine, pediatrics, emergency medicine, general surgery, obstetrics and gynecology, psychiatry, neurology, and 5 elective rotations), and 2 months for a thesis. The program has about 30 students per year and 4 clerkship sites for IM.

Program Description

The VABHS medicine clerkship hosts 1 to 2 PA students for 4-week blocks in the PAC unit of the CLC. Each student rotates on both PA and MD teams. Students follow 3 to 4 patients and participate fully in their care from admission to discharge; they prepare daily presentations and participate in medical management, family meetings, chart documentation, and care coordination with the interprofessional team. Students are provided a physical examination checklist and feedback form, and they are expected to track findings and record feedback and goals with their supervising preceptor weekly. They also make formal case presentations and participate in monthly medicine didactic rounds available to all VABHS IM students and trainees via videoconference.

In addition, beginning in July 2017, all PA students in the CLC began to participate in a 4-week Interprofessional Curriculum in Transitional Care. The curriculum includes 14 didactic lectures taught by 16 interprofessional faculty, including medicine, geriatric, and palliative care physicians; PAs; social workers; physical and occupational therapists; pharmacists; and a geriatric psychologist. The didactics include topics on the interprofessional team, the care continuum, teams and teamwork, interdisciplinary coordination of care, components of effective transitions in care, medication reconciliation, approaching difficult conversations, advance care planning, and quality improvement. The goal of the curriculum is to provide learners the knowledge, skills, and dispositions necessary for high-quality transitional care and interprofessional practice as well as specific training for effective and safe transfers of care between clinical settings. Although PA students are the main participants in this curriculum, all other learners in the PAC unit are also invited to attend the lectures.

The unique attributes of this training site include direct interaction with supervising PAs and physicians, rather than experiencing the traditional teaching hierarchy (with interns, residents, fellows); observation of the natural progression of disease of both acute care and primary care issues due to the longer length of stay (2 to 6 weeks, where the typical student will see the same patient 7 to 10 times during their rotation); exposure to a host of medically complex patients offering a multitude of clinical scenarios and abnormal physical exam findings; exposure to a hospice/palliative care ward and end-of-life care; and interaction within an interprofessional training environment of nursing, pharmacy, PT, OT, speech-language pathology, psychology, and SW trainees.

Program Evaluation

At the end of rotations continuously through the year, PA students electronically complete a site evaluation from the Boston University School of Medicine Physician Assistant Program. The evaluation consists of 14 questions: 6 about site quality and 8 about instruction quality. The questions are answered on a 5-point Likert scale. Also included are 2 open-ended response questions that ask what they liked about the rotation and what they felt could be improved. Results are anonymous, de-identified and blinded both to the program as well as the clerkship site. Results are aggregated and provided to program sites annually. Responses are converted to a dichotomous variable, where any good or excellent response (4 or 5) is considered positive and any neutral or below (3, 2, 1) is considered a nonpositive response.

Results

The clerkship site has been operational since June 22, 2015. There have been 59 students who participated in the rotation. A different scale in these evaluations was used between June 22, 2015, and September 13, 2015. Therefore, 7 responses were excluded from the analysis, leaving 52 usable evaluations. The responses were analyzed both in total (for the CLC as well as other IM rotation sites) and by individual clerkship year to look for any trends over time: September 14, 2015, through April 24, 2016; April 25, 2016, through April 28, 2017; and May 1, 2017, through March 1, 2018 (Table).

Site evaluations showed high satisfaction regarding the quality of the physical environment as well as the learning environment. Students endorsed the PAC unit having resources and physical space for them, such as a desk and computer, opportunity for participation in patient care, and parking (100%; n = 52). Site evaluations revealed high satisfaction with the quality of teaching and faculty encouragement and support of their learning (100%; n = 52). The evaluations revealed that bedside teaching was strong (94%; n = 49). The students reported high satisfaction with the volume of patients provided (92%; n = 48) as well as the diversity of diagnoses (92%; n = 48).

There were fewer positive responses in the first 2 years of the rotation with regard to formal lectures (50% and 67%; 7/14 and 16/24, respectively). In the third year of the rotation, students had a much higher satisfaction rate (93%; 13/14). This increased satisfaction was associated with the development and incorporation of the Interprofessional Curriculum in Transitional Care in 2017.

Discussion

Access to high-quality PA student clerkship sites has become a pressing issue in recent years because of increased competition for sites and a shortage of willing and well-trained preceptors. There has been marked growth in schools and enrollment across all medical professions. The Accreditation Review Commission on Education for the PA (ARC-PA) reported that the total number of accredited entry-level PA programs in 2018 was 246, with 58 new accredited programs projected by 2022.¹² The Joint Report of the 2013 Multi-Discipline Clerkship/Clinical Training Site Survey reported a 66% increase in first-year enrollment in PA programs from 2002 to 2012.⁵ Programs must implement alternative strategies to attract clinical sites (eg, academic appointments, increased clinical resources to training sites) or face continued challenges with recruiting training sites for their students. Postacute care may be a natural extension to expand the footprint for clinical sites for these programs, augmenting acute inpatient and outpatient rotations. This implementation would increase the pool of clinical training sites and preceptors.

The experience with this novel training site, based on PA student feedback and evaluations, has been positive, and the postacute setting can provide students with high-quality IM clinical experiences. Students report adequate patient volume and diversity. In addition, evaluations are comparable with that of other IM site rotations the students experience. Qualitative feedback has emphasized the value of following patients over longer periods; eg, weeks vs days (as in acute care) enabling students to build relationships with patients as well as observe a richer clinical spectrum of disease over a less compressed period. “Patients have complex issues, so from a medical standpoint it challenges you to think of new ways to manage their care,” commented a representative student. “It is really beneficial that you can follow them over time.”

Furthermore, in response to student feedback on didactics, an interprofessional curriculum was developed to add formal structure as well as to create a curriculum in care transitions. This curriculum provided a unique opportunity for PA students to receive formal instruction on areas of particular relevance for transitional care (eg, care continuum, end of life issues, and care transitions). The curriculum also allows the interprofessional faculty a unique and enjoyable opportunity for interprofessional collaboration.

The 1 month PAC rotation is augmented with inpatient IM and outpatient family medicine rotations, consequently giving exposure to the full continuum of care. The PAC setting provides learners multifaceted benefits: the opportunity to strengthen and develop the knowledge, attitudes, and skills necessary for IM; increased understanding of other professions by observing and interacting as a team caring for a patient over a longer period as opposed to the acute care setting; the ability to perform effective, efficient, and safe transfer between clinical settings; and broad exposure to transitional care. As a result, the PAC rotation enhances but does not replace the necessary and essential rotations of inpatient and outpatient medicine.

Moreover, this rotation provides unique and core IM training for PA students. Our site focuses on interprofessional collaboration, emphasizing the importance of team-based care, an essential concept in modern day medicine. Formal exposure to other care specialties, such as PT and OT, SW, and mental health, is essential for students to appreciate clinical medicine and a patient’s physical and mental experience over the course of a disease and clinical state. In addition, the physical exam checklist ensures that students are exposed to the full spectrum of IM examination findings during their rotation. Finally, weekly feedback forms require students to ask and receive concrete feedback from their supervising providers.

Limitations

The generalizability of this model requires careful consideration. VABHS is a tertiary care integrated health care system, enabling students to learn from patients moving through multiple care transitions in a single health care system. In addition, other settings may not have the staffing or clinical volume to sustain such a model. All PAC clinical faculty teach voluntarily, and local leadership has set expectations for all clinicians to participate in teaching of trainees and PA students. Evaluations also note less diversity in the patient population, a challenge that some VA facilities face. This issue could be addressed by ensuring that students also have IM rotations at other inpatient medical facilities. A more balanced experience, where students reap the positive benefits of PAC but do not lose exposure to a diverse patient pool, could result. Furthermore, some of the perceived positive impacts also may be related to professional and personal attributes of the teaching clinicians rather than to the PAC setting.

Conclusion

PAC settings can be effective training sites for medicine clerkships for PA students and can provide high-quality training in IM as PA programs continue to expand. This setting offers students exposure to interprofessional, team-based care and the opportunity to care for patients with a broad range of disease complexity. Learning is further enhanced by the ability to follow patients longitudinally over their disease course as well as to work directly with teaching faculty and other interprofessional health care professionals. Evaluations of this novel clerkship experience have shown high levels of student satisfaction in knowledge growth, clinical skills, bedside teaching, and mentorship.

Acknowledgments
We thank Juman Hijab for her critical role in establishing and maintaining the clerkship. We thank Steven Simon, Matt Russell, and Thomas Parrino for their leadership and guidance in establishing and maintaining the clerkship. We thank the Boston University School of Medicine Physician Assistant Program Director Mary Warner for her support and guidance in creating and supporting the clerkship. In addition, we thank the interprofessional education faculty for their dedicated involvement in teaching, including Stephanie Saunders, Lindsay Lefers, Jessica Rawlins, Lindsay Brennan, Angela Viani, Eric Charette, Nicole O’Neil, Susan Nathan, Jordana Meyerson, Shivani Jindal, Wei Shen, Amy Hanson, Gilda Cain, and Kate Hinrichs.

*This version has been corrected. In the original version the first sentence incorrectly referred to moral injury instead of burnout.

For more than a decade, the term burnout has been used to describe clinician distress.^1,2 Although some clinicians in federal health care systems may be protected from some of the drivers of burnout, other federal practitioners suffer from rule-driven health care practices and distant, top-down administration. The demand for health care is expanding, driven by the aging of the US population.³ Massive information technology investments, which promised efficiency for health care providers,⁴ have instead delivered a triple blow: They have diverted capital resources that might have been used to hire additional caregivers,⁵ diverted the time and attention of those already engaged in patient care,⁶ and done little to improve patient outcomes.⁷ Reimbursements are falling, and the only way for health systems to maintain their revenue is to increase the number of patients each clinician sees per day.⁸ As the resources of time and attention shrink, and as spending continues with no improvement in patient outcomes, clinician distress is on the rise.⁹ It will be important to understand exactly what the drivers of the problem are for federal clinicians so that solutions can be appropriately targeted. The first step in addressing the epidemic of physician distress is using the most accurate terminology to describe it.

Freudenberger defined burnout in 1975 as a constellation of symptoms—malaise, fatigue, frustration, cynicism, and inefficacy—that arise from “making excessive demands on energy, strength, or resources” in the workplace.¹⁰ The term was borrowed from other fields and applied to health care in the hopes of readily transferring the solutions that had worked in other industries to address a growing crisis among physicians. Unfortunately, the crisis in health care has proven resistant to solutions that have worked elsewhere, and many clinicians have resisted being characterized as burned out, citing a subtle, elusive disconnect between what they have experienced and what burnout encapsulates.

In July 2018, the conversation about clinician distress shifted with an article we wrote in STAT that described the moral injury of health care.¹¹ The concept of moral injury was first described in service members who returned from the Vietnam War with symptoms that loosely fit a diagnosis of posttraumatic stress disorder (PTSD), but which did not respond to standard PTSD treatment and contained symptoms outside the PTSD constellation.¹² On closer assessment, what these service members were experiencing had a different driver. Whereas those with PTSD experienced a real and imminent threat to their mortality and had come back deeply concerned for their individual, physical safety, those with this different presentation experienced repeated insults to their morality and had returned questioning whether they were still, at their core, moral beings. They had been forced, in some way, to act contrary to what their beliefs dictated was right by killing civilians on orders from their superiors, for example. This was a different category of psychological injury that required different treatment.

Moral injury occurs when we perpetrate, bear witness to, or fail to prevent an act that transgresses our deeply held moral beliefs. In the health care context, that deeply held moral belief is the oath each of us took when embarking on our paths as health care providers: Put the needs of patients first. That oath is the lynchpin of our working lives and our guiding principle when searching for the right course of action. But as clinicians, we are increasingly forced to consider the demands of other stakeholders—the electronic medical record (EMR), the insurers, the hospital, the health care system, even our own financial security—before the needs of our patients. Every time we are forced to make a decision that contravenes our patients’ best interests, we feel a sting of moral injustice. Over time, these repetitive insults amass into moral injury.

The difference between burnout and moral injury is important because using different terminology reframes the problem and the solutions. Burnout suggests that the problem resides within the individual, who is in some way deficient. It implies that the individual lacks the resources or resilience to withstand the work environment. Since the problem is in the individual, the solutions to burnout must be in the individual, too, and therefore, it is the individual’s responsibility to find and implement them. Many of the solutions to physician distress posited to date revolve around this conception; hence, the focus on yoga, mindfulness, wellness retreats, and meditation.¹³ While there is nothing inherently wrong with any of those practices, it is absurd to believe that yoga will solve the problems of treating a cancer patient with a declined preauthorization for chemotherapy, having no time to discuss a complex diagnosis, or relying on a computer system that places metrics ahead of communication. These problems are not the result of some failing on the part of the individual clinician.

Moral injury, on the other hand, describes the challenge of simultaneously knowing what care patients need but being unable to provide it due to constraints that are beyond our control. Moral injury is the consequence of the ever-present double binds in health care: Do we take care of our patient, the hospital, the insurer, the EMR, the health care system, or our productivity metrics first? There should be only 1 answer to that question, but the current business framework of medicine pressures us to serve all these masters at once. Moral injury locates the source of distress in a broken system, not a broken individual, and allows us to direct solutions at the causes of distress. And in the end, addressing the drivers of moral injury on a large scale may be the most effective preventive treatment for its cumulative effects among health care providers.

The long-term solutions to moral injury demand changes in the business framework of health care. The solutions reside not in promoting mindfulness or resilience among individual physicians, but in creating a health care environment that finally acknowledges the value of the time clinicians and patients spend together developing the trust, understanding, and compassion that accompany a true relationship. The long-term solutions to moral injury include a health care system that prioritizes healing over profit and that trusts its clinicians to always put their patients’ best interests first.

Treating moral injury will not be simple. It cannot happen quickly, and it will not happen without widespread clinician engagement. Change can begin when clinicians identify the double binds they face every day and convey those challenges to their administrators. If administrators and clinicians are willing to work together to resolve these double binds, health care will improve for everyone.

The following are our recommendations for how you can bring change both locally and on a broader scale.

Bring together the 2 sides of the health care house: administrators and clinicians. Invite administrators to join you on rounds, in clinic, or in the operating room. Ask them to follow you during a night of call or to spend an overnight shift with you in the emergency department. The majority of people, including health care administrators, have had only glancing encounters with the medical system. They see their primary care doctor, have regular screening procedures, and maybe get treated for a routine illness or injury. None of those encounters expose them to the depth of challenge in the system.

It takes exposure over a longer duration, or with greater intensity, to appreciate the tensions and double binds that patients and clinicians face regularly.^14,15 Whether or not the administrators accept your invitation, you must also ask to see the challenges from their side. Block out an afternoon, a day, or a week to follow them and learn where they struggle in their work. Only when we understand the other party’s perspective can we truly begin to empathize and communicate meaningfully. That profound understanding is the place where commonality and compromises are found.

Make clinician satisfaction a financial priority. Although care team well-being is now part of the quadruple aim (patient experience, population health, reducing costs, and provider experience), organizations must be held accountable to ensure it is a priority. If we choose to link patient satisfaction with clinician compensation, why not link clinician satisfaction with executive compensation?

Make sure every physician leader has and uses the cell phone number of his or her legislators. Hospitals and big pharma have nearly bottomless lobbying budgets, which makes competing with them for lawmakers’ attention a formidable prospect. Despite this, physician leaders (ie, chief wellness officer, department chairperson, medical society president, etc) have a responsibility to communicate with legislators about the needs of patients (their constituents) and what role our legislators can play in fulfilling those needs. We must understand how policy, regulation, and legislation work, and we need to find seats at every table where the decisions that impact clinical care are made. The first step is opening lines of communication with those who have the power to enact large-scale change.

Reestablish a sense of community among clinicians. Too often clinicians are pitted against one another as resources shrink. Doctors compete with each other for referrals, advanced practitioners and nurses compete with doctors, and everyone feels overstressed. What we tend to forget is that we are all working toward the same goal: To give patients the best care possible. It’s time to view each other with the presumption of charity and to have each other’s backs. Uniting for support, camaraderie, mentorship, and activism is a necessary step in making change.

*This version has been corrected. In the original version the first sentence incorrectly referred to moral injury instead of burnout.

For more than a decade, the term burnout has been used to describe clinician distress.^1,2 Although some clinicians in federal health care systems may be protected from some of the drivers of burnout, other federal practitioners suffer from rule-driven health care practices and distant, top-down administration. The demand for health care is expanding, driven by the aging of the US population.³ Massive information technology investments, which promised efficiency for health care providers,⁴ have instead delivered a triple blow: They have diverted capital resources that might have been used to hire additional caregivers,⁵ diverted the time and attention of those already engaged in patient care,⁶ and done little to improve patient outcomes.⁷ Reimbursements are falling, and the only way for health systems to maintain their revenue is to increase the number of patients each clinician sees per day.⁸ As the resources of time and attention shrink, and as spending continues with no improvement in patient outcomes, clinician distress is on the rise.⁹ It will be important to understand exactly what the drivers of the problem are for federal clinicians so that solutions can be appropriately targeted. The first step in addressing the epidemic of physician distress is using the most accurate terminology to describe it.

Freudenberger defined burnout in 1975 as a constellation of symptoms—malaise, fatigue, frustration, cynicism, and inefficacy—that arise from “making excessive demands on energy, strength, or resources” in the workplace.¹⁰ The term was borrowed from other fields and applied to health care in the hopes of readily transferring the solutions that had worked in other industries to address a growing crisis among physicians. Unfortunately, the crisis in health care has proven resistant to solutions that have worked elsewhere, and many clinicians have resisted being characterized as burned out, citing a subtle, elusive disconnect between what they have experienced and what burnout encapsulates.

In July 2018, the conversation about clinician distress shifted with an article we wrote in STAT that described the moral injury of health care.¹¹ The concept of moral injury was first described in service members who returned from the Vietnam War with symptoms that loosely fit a diagnosis of posttraumatic stress disorder (PTSD), but which did not respond to standard PTSD treatment and contained symptoms outside the PTSD constellation.¹² On closer assessment, what these service members were experiencing had a different driver. Whereas those with PTSD experienced a real and imminent threat to their mortality and had come back deeply concerned for their individual, physical safety, those with this different presentation experienced repeated insults to their morality and had returned questioning whether they were still, at their core, moral beings. They had been forced, in some way, to act contrary to what their beliefs dictated was right by killing civilians on orders from their superiors, for example. This was a different category of psychological injury that required different treatment.

Moral injury occurs when we perpetrate, bear witness to, or fail to prevent an act that transgresses our deeply held moral beliefs. In the health care context, that deeply held moral belief is the oath each of us took when embarking on our paths as health care providers: Put the needs of patients first. That oath is the lynchpin of our working lives and our guiding principle when searching for the right course of action. But as clinicians, we are increasingly forced to consider the demands of other stakeholders—the electronic medical record (EMR), the insurers, the hospital, the health care system, even our own financial security—before the needs of our patients. Every time we are forced to make a decision that contravenes our patients’ best interests, we feel a sting of moral injustice. Over time, these repetitive insults amass into moral injury.

The difference between burnout and moral injury is important because using different terminology reframes the problem and the solutions. Burnout suggests that the problem resides within the individual, who is in some way deficient. It implies that the individual lacks the resources or resilience to withstand the work environment. Since the problem is in the individual, the solutions to burnout must be in the individual, too, and therefore, it is the individual’s responsibility to find and implement them. Many of the solutions to physician distress posited to date revolve around this conception; hence, the focus on yoga, mindfulness, wellness retreats, and meditation.¹³ While there is nothing inherently wrong with any of those practices, it is absurd to believe that yoga will solve the problems of treating a cancer patient with a declined preauthorization for chemotherapy, having no time to discuss a complex diagnosis, or relying on a computer system that places metrics ahead of communication. These problems are not the result of some failing on the part of the individual clinician.

Moral injury, on the other hand, describes the challenge of simultaneously knowing what care patients need but being unable to provide it due to constraints that are beyond our control. Moral injury is the consequence of the ever-present double binds in health care: Do we take care of our patient, the hospital, the insurer, the EMR, the health care system, or our productivity metrics first? There should be only 1 answer to that question, but the current business framework of medicine pressures us to serve all these masters at once. Moral injury locates the source of distress in a broken system, not a broken individual, and allows us to direct solutions at the causes of distress. And in the end, addressing the drivers of moral injury on a large scale may be the most effective preventive treatment for its cumulative effects among health care providers.

The long-term solutions to moral injury demand changes in the business framework of health care. The solutions reside not in promoting mindfulness or resilience among individual physicians, but in creating a health care environment that finally acknowledges the value of the time clinicians and patients spend together developing the trust, understanding, and compassion that accompany a true relationship. The long-term solutions to moral injury include a health care system that prioritizes healing over profit and that trusts its clinicians to always put their patients’ best interests first.

Treating moral injury will not be simple. It cannot happen quickly, and it will not happen without widespread clinician engagement. Change can begin when clinicians identify the double binds they face every day and convey those challenges to their administrators. If administrators and clinicians are willing to work together to resolve these double binds, health care will improve for everyone.

The following are our recommendations for how you can bring change both locally and on a broader scale.

Bring together the 2 sides of the health care house: administrators and clinicians. Invite administrators to join you on rounds, in clinic, or in the operating room. Ask them to follow you during a night of call or to spend an overnight shift with you in the emergency department. The majority of people, including health care administrators, have had only glancing encounters with the medical system. They see their primary care doctor, have regular screening procedures, and maybe get treated for a routine illness or injury. None of those encounters expose them to the depth of challenge in the system.

It takes exposure over a longer duration, or with greater intensity, to appreciate the tensions and double binds that patients and clinicians face regularly.^14,15 Whether or not the administrators accept your invitation, you must also ask to see the challenges from their side. Block out an afternoon, a day, or a week to follow them and learn where they struggle in their work. Only when we understand the other party’s perspective can we truly begin to empathize and communicate meaningfully. That profound understanding is the place where commonality and compromises are found.

Make clinician satisfaction a financial priority. Although care team well-being is now part of the quadruple aim (patient experience, population health, reducing costs, and provider experience), organizations must be held accountable to ensure it is a priority. If we choose to link patient satisfaction with clinician compensation, why not link clinician satisfaction with executive compensation?

Make sure every physician leader has and uses the cell phone number of his or her legislators. Hospitals and big pharma have nearly bottomless lobbying budgets, which makes competing with them for lawmakers’ attention a formidable prospect. Despite this, physician leaders (ie, chief wellness officer, department chairperson, medical society president, etc) have a responsibility to communicate with legislators about the needs of patients (their constituents) and what role our legislators can play in fulfilling those needs. We must understand how policy, regulation, and legislation work, and we need to find seats at every table where the decisions that impact clinical care are made. The first step is opening lines of communication with those who have the power to enact large-scale change.

Reestablish a sense of community among clinicians. Too often clinicians are pitted against one another as resources shrink. Doctors compete with each other for referrals, advanced practitioners and nurses compete with doctors, and everyone feels overstressed. What we tend to forget is that we are all working toward the same goal: To give patients the best care possible. It’s time to view each other with the presumption of charity and to have each other’s backs. Uniting for support, camaraderie, mentorship, and activism is a necessary step in making change.

*This version has been corrected. In the original version the first sentence incorrectly referred to moral injury instead of burnout.

For more than a decade, the term burnout has been used to describe clinician distress.^1,2 Although some clinicians in federal health care systems may be protected from some of the drivers of burnout, other federal practitioners suffer from rule-driven health care practices and distant, top-down administration. The demand for health care is expanding, driven by the aging of the US population.³ Massive information technology investments, which promised efficiency for health care providers,⁴ have instead delivered a triple blow: They have diverted capital resources that might have been used to hire additional caregivers,⁵ diverted the time and attention of those already engaged in patient care,⁶ and done little to improve patient outcomes.⁷ Reimbursements are falling, and the only way for health systems to maintain their revenue is to increase the number of patients each clinician sees per day.⁸ As the resources of time and attention shrink, and as spending continues with no improvement in patient outcomes, clinician distress is on the rise.⁹ It will be important to understand exactly what the drivers of the problem are for federal clinicians so that solutions can be appropriately targeted. The first step in addressing the epidemic of physician distress is using the most accurate terminology to describe it.

Freudenberger defined burnout in 1975 as a constellation of symptoms—malaise, fatigue, frustration, cynicism, and inefficacy—that arise from “making excessive demands on energy, strength, or resources” in the workplace.¹⁰ The term was borrowed from other fields and applied to health care in the hopes of readily transferring the solutions that had worked in other industries to address a growing crisis among physicians. Unfortunately, the crisis in health care has proven resistant to solutions that have worked elsewhere, and many clinicians have resisted being characterized as burned out, citing a subtle, elusive disconnect between what they have experienced and what burnout encapsulates.

In July 2018, the conversation about clinician distress shifted with an article we wrote in STAT that described the moral injury of health care.¹¹ The concept of moral injury was first described in service members who returned from the Vietnam War with symptoms that loosely fit a diagnosis of posttraumatic stress disorder (PTSD), but which did not respond to standard PTSD treatment and contained symptoms outside the PTSD constellation.¹² On closer assessment, what these service members were experiencing had a different driver. Whereas those with PTSD experienced a real and imminent threat to their mortality and had come back deeply concerned for their individual, physical safety, those with this different presentation experienced repeated insults to their morality and had returned questioning whether they were still, at their core, moral beings. They had been forced, in some way, to act contrary to what their beliefs dictated was right by killing civilians on orders from their superiors, for example. This was a different category of psychological injury that required different treatment.

Moral injury occurs when we perpetrate, bear witness to, or fail to prevent an act that transgresses our deeply held moral beliefs. In the health care context, that deeply held moral belief is the oath each of us took when embarking on our paths as health care providers: Put the needs of patients first. That oath is the lynchpin of our working lives and our guiding principle when searching for the right course of action. But as clinicians, we are increasingly forced to consider the demands of other stakeholders—the electronic medical record (EMR), the insurers, the hospital, the health care system, even our own financial security—before the needs of our patients. Every time we are forced to make a decision that contravenes our patients’ best interests, we feel a sting of moral injustice. Over time, these repetitive insults amass into moral injury.

The difference between burnout and moral injury is important because using different terminology reframes the problem and the solutions. Burnout suggests that the problem resides within the individual, who is in some way deficient. It implies that the individual lacks the resources or resilience to withstand the work environment. Since the problem is in the individual, the solutions to burnout must be in the individual, too, and therefore, it is the individual’s responsibility to find and implement them. Many of the solutions to physician distress posited to date revolve around this conception; hence, the focus on yoga, mindfulness, wellness retreats, and meditation.¹³ While there is nothing inherently wrong with any of those practices, it is absurd to believe that yoga will solve the problems of treating a cancer patient with a declined preauthorization for chemotherapy, having no time to discuss a complex diagnosis, or relying on a computer system that places metrics ahead of communication. These problems are not the result of some failing on the part of the individual clinician.

Moral injury, on the other hand, describes the challenge of simultaneously knowing what care patients need but being unable to provide it due to constraints that are beyond our control. Moral injury is the consequence of the ever-present double binds in health care: Do we take care of our patient, the hospital, the insurer, the EMR, the health care system, or our productivity metrics first? There should be only 1 answer to that question, but the current business framework of medicine pressures us to serve all these masters at once. Moral injury locates the source of distress in a broken system, not a broken individual, and allows us to direct solutions at the causes of distress. And in the end, addressing the drivers of moral injury on a large scale may be the most effective preventive treatment for its cumulative effects among health care providers.

The long-term solutions to moral injury demand changes in the business framework of health care. The solutions reside not in promoting mindfulness or resilience among individual physicians, but in creating a health care environment that finally acknowledges the value of the time clinicians and patients spend together developing the trust, understanding, and compassion that accompany a true relationship. The long-term solutions to moral injury include a health care system that prioritizes healing over profit and that trusts its clinicians to always put their patients’ best interests first.

Treating moral injury will not be simple. It cannot happen quickly, and it will not happen without widespread clinician engagement. Change can begin when clinicians identify the double binds they face every day and convey those challenges to their administrators. If administrators and clinicians are willing to work together to resolve these double binds, health care will improve for everyone.

The following are our recommendations for how you can bring change both locally and on a broader scale.

Bring together the 2 sides of the health care house: administrators and clinicians. Invite administrators to join you on rounds, in clinic, or in the operating room. Ask them to follow you during a night of call or to spend an overnight shift with you in the emergency department. The majority of people, including health care administrators, have had only glancing encounters with the medical system. They see their primary care doctor, have regular screening procedures, and maybe get treated for a routine illness or injury. None of those encounters expose them to the depth of challenge in the system.

It takes exposure over a longer duration, or with greater intensity, to appreciate the tensions and double binds that patients and clinicians face regularly.^14,15 Whether or not the administrators accept your invitation, you must also ask to see the challenges from their side. Block out an afternoon, a day, or a week to follow them and learn where they struggle in their work. Only when we understand the other party’s perspective can we truly begin to empathize and communicate meaningfully. That profound understanding is the place where commonality and compromises are found.

Make clinician satisfaction a financial priority. Although care team well-being is now part of the quadruple aim (patient experience, population health, reducing costs, and provider experience), organizations must be held accountable to ensure it is a priority. If we choose to link patient satisfaction with clinician compensation, why not link clinician satisfaction with executive compensation?

Make sure every physician leader has and uses the cell phone number of his or her legislators. Hospitals and big pharma have nearly bottomless lobbying budgets, which makes competing with them for lawmakers’ attention a formidable prospect. Despite this, physician leaders (ie, chief wellness officer, department chairperson, medical society president, etc) have a responsibility to communicate with legislators about the needs of patients (their constituents) and what role our legislators can play in fulfilling those needs. We must understand how policy, regulation, and legislation work, and we need to find seats at every table where the decisions that impact clinical care are made. The first step is opening lines of communication with those who have the power to enact large-scale change.

Reestablish a sense of community among clinicians. Too often clinicians are pitted against one another as resources shrink. Doctors compete with each other for referrals, advanced practitioners and nurses compete with doctors, and everyone feels overstressed. What we tend to forget is that we are all working toward the same goal: To give patients the best care possible. It’s time to view each other with the presumption of charity and to have each other’s backs. Uniting for support, camaraderie, mentorship, and activism is a necessary step in making change.