Faculty Bibliography

INTRODUCTION: Goal-directed echocardiography (GDE) is used to answer specific clinical questions that provide invaluable information to physicians managing a hemodynamically unstable patient. We studied perception and ability of house staff previously trained in GDE to accurately diagnose common causes of cardiac arrest during simulated advanced cardiac life support (ACLS); we compared their results with those of expert echocardiographers. METHODS: Eleven pulmonary and critical care medicine fellows, 7 emergency medicine residents, and 5 cardiologists board certified in echocardiography were enrolled. Baseline ability to acquire 4 transthoracic echocardiography views was assessed, and participants were exposed to 6 simulated cardiac arrests and were asked to perform a GDE during ACLS. House staff performance was compared with the performance of 5 expert echocardiographers. RESULTS: Average baseline and scenario views by house staff were of good or excellent quality 89% and 83% of the time, respectively. Expert average baseline and scenario views were always of good or excellent quality. House staff and experts made the correct diagnosis in 68% and 77% of cases, respectively. On average, participants required 1.5 pulse checks to make the correct diagnosis. Of house staff, 94% perceived this study as an accurate assessment of ability. CONCLUSIONS: In an ACLS-compliant manner, house staff are capable of diagnosing management-altering pathologies the majority of the time, and they reach similar diagnostic conclusions in the same amount of time as expert echocardiographers in a simulated cardiac arrest scenario.

OBJECTIVE: Several studies have questioned the effectiveness of rapid-response systems when measured by outcomes such as decreased overall hospital mortality or cardiac arrest rates. We studied an alternative outcome of rapid-response system implementation, namely, its effect on goals of care and designation of do not resuscitate. DESIGN: Retrospective chart review. SETTING: Veterans Administration Hospital in New York City. SUBJECTS: All patients requiring a medical emergency team call. INTERVENTIONS: None MEASUREMENTS AND MAIN RESULTS: : Monthly hospital census and discharge data, death occurrences, and do-not-resuscitate order placements were collected over an 8-year pre-medical emergency team and 5-year post-medical emergency team period. All medical emergency team calls and subsequent transfers to a critical care unit were reviewed and correlated to the placement and timing of do-not-resuscitate orders. Interrupted time-series analysis was used to evaluate the impact of the medical emergency team implementation on the change in trend of do-not-resuscitate orders and the hospital mortality. A total of 390 medical emergency team calls were associated with 109 do-not-resuscitate orders (28%). Of the 209 medical emergency team calls (54%) resulting in transfer to a critical care unit, 66 were associated with do-not-resuscitate orders, 73% of which were obtained after transfer. The odds of becoming do not resuscitate for a patient going to the ICU after the medical emergency team call were 2.9 (95% CI, 1.6-5.5; p = 0.001) times greater than for patients staying on the floors after the medical emergency team call. The medical emergency team implementation significantly changed the trend of do-not-resuscitate orders (p < 0.001) but had no impact on hospital mortality rate (p = 0.638). CONCLUSION: Implementation of a rapid-response system was associated with an increase in do-not-resuscitate order placement. As a sentinel event, medical emergency team activation and transfer to a critical care unit foster consideration of goals of care and frequently results in a transition to a palliative care strategy.

PURPOSE: Goal-directed echocardiography (GDE) is used to answer specific clinical questions which can provide invaluable and timely information to the critical care physician. Studies that assess competency are lacking. We studied perception and ability of housestaff previously trained in GDE to accurately diagnose common causes of cardiac arrest during simulated cardiac arrest scenarios, and we compared them to expert echocardiographers. METHODS: 14 housestaff subjects with prior GDE training were enrolled. Subjects answered a pre-study questionnaire and had time to familiarize themselves with the transthoracic echocardiography simulator. A baseline assessment was conducted whereby subjects obtained four standard cardiac windows (parasternal long, parasternal short, subcostal, and apical four chamber). Subjects were exposed to six simulated cardiac arrest scenarios. They were given relevant clinical information and were asked to perform a GDE during pulse checks which lasted ten seconds. Three GDE attempts were allowed and if no diagnosis was offered, a final twenty second interval was allowed. Subjects were debriefed and filled out a post-study questionnaire. All echocardiography views were graded on a scale of zero to three. Subject performance was compared to the performance of three expert echocardiographers. RESULTS: 21% of subjects reported comfort using GDE independently, while 71% preferred attending oversight. Baseline and scenario views by subjects were of good quality 93% and 79% of the time, respectively. Expert baseline and scenario views were of good quality 100% of the time. Subjects and experts made the correct diagnosis in 68% and 72% of cases, respectively. On average, subjects and experts required 1.5 pulse checks for the correct diagnosis. 93% of subjects perceived this study as an accurate assessment of ability and felt more comfortable with GDE at its conclusion. CONCLUSIONS: Housestaff with prior GDE training reach similar diagnostic conclusions in the same amount of tim!

INTRODUCTION: Hodgkinas Lymphoma (HL) is commonly associated with peripheral eosinophilia but not eosiniophilic organ infiltration. We report a case of an HL patient suffering two discrete episodes of Acute Eosinophilic Pneumonia (AEP). CASE PRESENTATION: A 34 year-old man with refractory HL (to chemotherapies and bone-marrow transplant, currently on Lenalinomide), intermittent eosinophilia and AEP (17 months prior due to Gemcitabine) presented with complaints of fever, dyspnea and productive cough. On exam he was febrile, tachycardic, tachypneic and hypoxemic. Radiography revealed a left upper lobe infiltrate. Labwork revealed eosiniophilia. He was admitted to Hematology and placed on broad-spectrum antibiotics but within hours developed hypoxemic respiratory failure requiring MICU transfer. Due to his pulmonary infiltrates and eosinophilia he underwent emergent bronchoscopy; BAL revealed 31% eosinophils without organisms on cytology. All initial tests for bacteria, fungi, AFB and Ova/Parasites were negative, a diagnosis of AEP was made and IV solumedrol initiated. Over the following day his respiratory status dramatically improved, his fever resolved, he was weaned off supplementary oxygen and transferred to Hematology. He remained afebrile as antibiotics were narrowed to a macrolide and steroids transitioned to oral taper. All final cultures/serologies returned negative. After discharge he remained asymptomatic off Lenalidomide; repeat imaging confirmed resolution of the infiltrate. DISCUSSION: AEP is characterized by acute-onset fever, cough, dyspnea and pulmonary infiltrates which if untreated can progress to respiratory failure; the etiology is likely a hypersensitivity reaction to novel antigens. Diagnosis requires BAL with over 25% eosinophils and is a diagnosis of exclusion; treatment includes elimination of the offending agent and steroids (usually with rapid recovery). Lenalidomide, a derivative of thalidomide with immuno-modulatory and anti-angiogenic properties, is FDA approved for!

Rationale Current methods available to assess rapid response management skills are subjective and variable. High-fidelity simulation (HFS), however, allows for objective and standardized measurement of these skills. Physicians-in-training can undergo the same simulation and be scored against one another to determine their level of proficiency when leading a rapid response. We created a critical care simulation scenario and corresponding checklist assessment tool to measure specific behaviors related to effective management of a rapid response. Methods Forty-four second-year internal medicine residents underwent HFS of a critically-ill patient with pulseless electrical activity. The simulation was designed to elicit leadership qualities, communication skills, decision-making and resource management during a rapid response. A comprehensive checklist was developed from expert consultation that itemized behaviors as having been "well done," "partially done," or "not done." Video-recordings of simulations were reviewed by two independent raters, each of which underwent extensive training in checklist use prior to project initiation. Results Composite scores from both raters revealed that 39% (103/264) and 55% (243/440) of resident scores were "well done" for leadership and communication skills during the rapid response, respectively. Twenty-six percent (46/176) of scores were "well done" for decision-making and 42% (37/88) of scores were "well done" for resource utilization skills. Cronbach's Alpha analysis of internal consistency of composite "well done" scores was good, with a value of 0.869. Conclusions HFS with use of a corresponding checklist assessment tool is an objective and effective way in which to measure rapid response management skills. During our simulation, a minority of residents received "well done" scores for leadership, decision-making and resource utilization. This may reflect residents' discomfort when placed in a supervisory role during rapid responses, a paucity of exposure to such situations, or inadequate training in running rapid responses. In order to better educate physicians-in-training, a standardized and objective measurement tool must be used. Our checklist assessment tool demonstrates good internal consistency. Therefore, it can be used to objectively measure behaviors that exemplify organization and management of rapid responses

PURPOSE: Current methods of measuring medical professionalism are subjective. They often do not account for high-pressure environments where professionalism can be strained, such as the intensive care unit (ICU). High-fidelity simulation (HFS) is a technology in which standardized, high-pressure patient scenarios are practiced. The standardization of HFS allows for objective measurement of professionalism-specific behaviors found in the ICU. METHODS: Forty-four second-year internal medicine residents underwent HFS of a critically-ill patient. Professionalism-specific behaviors pertaining to obtaining informed consent for central venous line (CVL) placement and disclosure of an iatrogenic complication secondary to the CVL were evaluated. A comprehensive checklist itemized pre-determined professionalism-specific behaviors as having been 'well done, ' 'partially done, ' or 'not done. ' The checklist was formulated from expert consultation, and items included objectively defined behaviors and excluded medical knowledge prerequisites. Video-recordings of simulations were reviewed by three independent raters, each of which underwent training prior to project initiation. RESULTS: Composite scores from three raters revealed that 74.2% (98/132) and 67.4% (89/132) of residents received 'well done ' scores for discussion of CVL benefits and risks, respectively. Sixteen percent (21/132) of residents received 'well done ' scores regarding discussion of alternatives to CVL placement. Concerning disclosure of the iatrogenic complication, 22% (29/132) of residents performed this task well. Cronbach's alpha analyses of internal consistency were .813 for obtaining informed consent and .709 for disclosure of the iatrogenic complication. CONCLUSIONS: During our ICU simulation, a majority of residents outlined benefits and risks to CVL placement well, though most neglected to discuss alternatives. Furthermore, a minority of residents disclosed the etiology of the iatrogenic complication. It is possible that the high-stress simulated ICU environment contributed to the low prevalence of these behaviors. CLINICAL IMPLICATIONS: Objective tools to evaluate medical professionalism are scarce. However, our assessment tool checklist demonstrates good internal consistency, and therefore HFS of a critically-ill patient can be used to reliably and objectively measure pre-determined professionalism-specific behaviors