Albuterol has been shown to improve outcomes in patients presenting with bronchospasm in the prehospital setting. However, the administration of albuterol therapy is restricted by scope of practice to Advanced Life Support (ALS) in many Emergency Medical Services (EMS) systems throughout the country. Basic Life Support (BLS) providers have traditionally only assisted patients with using their own albuterol, but have not been granted authority to identify patients with bronchospasm and administer albuterol to patients who may benefit from bronchodilator therapy.

Study ObjectivesRecent evidence suggests patients are unable to distinguish severity of illness or appropriateness of emergency department (ED) visits based on their presenting complaints. Many patients seek access to outpatient care prior to their ED visit, but there is little evidence to suggest that doing so leads to a more appropriate or "informed" triage decision. A marker of appropriateness of ED visits may be admission rate. This study will determine whether admission rates differ between patients who self-triage to the ED compared to those who are sent there by a health care provider, and whether there is a difference based on the type of outpatient provider contact patients experienced.MethodsA prospective cross-sectional study of adult ED patients who presented to a single tertiary care referral Level I trauma center with 115,000 annual ED visits. Consenting patients in the ED were surveyed during regular intervals during days and evenings 7 days/week. The survey was pilot-tested and validated in the target population. Patients were asked whether they had attempted to contact an outside provider prior to their ED visit, and if successful, what type of provider they contacted, whether it was in person or by phone, and what instructions they received. Responses were then matched with ED disposition data. Those patients physically seen in an office or medical aid unit, or told by phone to go the ED were considered "sent" to the ED. They were compared with the group that "self-triaged." Patients were considered to have had an "informed" decision to be sent to the ED if they were seen in person or spoke with a doctor by phone. Pearson's Chi-Square testing was used to determine associations between the type of outpatient provider contact and ED disposition.ResultsThere was no difference in admission rate between those who attempted to contact their doctors (98/207, 47%) and those who did not (122/151, 48%). There was no difference between those sent to the ED (86/166, 51%) and those who self-triaged (134/292, 46%). However, amongst those who were sent to the ED after phone contact, there was a higher admission rate in those who had spoken to a doctor (25/38, 66%) as compared to those who had spoken to a non-physician such as a receptionist, nurse, or midlevel provider (23/61, 38%), p<0.01. Patients who received an informed decision to be sent to the ED were more likely to be admitted (59/99, 60%) than others who were sent to the ED (23/61, 38%), p<0.01.ConclusionsPatients who seek outpatient care prior to their ED visit are admitted to the hospital just as often as those who self-triage. This supports recent studies demonstrating the inability of patients to determine the appropriateness of an ED visit. Some patients may be sent to an ED for testing or interventions not otherwise available as an outpatient. But, if admission rate is an indicator of appropriateness of ED visits, then a more informed decision to send the patient to the ED may reduce avoidable ED utilization. Patients who are physically seen by a provider, or at least speak to a physician by phone, are more likely to be admitted, compared with those who are directed to the ED by a non-physician over the phone. This suggests that improving patients' access to a physician, at least by phone, or expanding acute unscheduled care options for both assessments and advanced testing, may reduce ED visits that do not require hospital admission. Study ObjectivesRecent evidence suggests patients are unable to distinguish severity of illness or appropriateness of emergency department (ED) visits based on their presenting complaints. Many patients seek access to outpatient care prior to their ED visit, but there is little evidence to suggest that doing so leads to a more appropriate or "informed" triage decision. A marker of appropriateness of ED visits may be admission rate. This study will determine whether admission rates differ between patients who self-triage to the ED compared to those who are sent there by a health care provider, and whether there is a difference based on the type of outpatient provider contact patients experienced. Recent evidence suggests patients are unable to distinguish severity of illness or appropriateness of emergency department (ED) visits based on their presenting complaints. Many patients seek access to outpatient care prior to their ED visit, but there is little evidence to suggest that doing so leads to a more appropriate or "informed" triage decision. A marker of appropriateness of ED visits may be admission rate. This study will determine whether admission rates differ between patients who self-triage to the ED compared to those who are sent there by a health care provider, and whether there is a difference based on the type of outpatient provider contact patients experienced. MethodsA prospective cross-sectional study of adult ED patients who presented to a single tertiary care referral Level I trauma center with 115,000 annual ED visits. Consenting patients in the ED were surveyed during regular intervals during days and evenings 7 days/week. The survey was pilot-tested and validated in the target population. Patients were asked whether they had attempted to contact an outside provider prior to their ED visit, and if successful, what type of provider they contacted, whether it was in person or by phone, and what instructions they received. Responses were then matched with ED disposition data. Those patients physically seen in an office or medical aid unit, or told by phone to go the ED were considered "sent" to the ED. They were compared with the group that "self-triaged." Patients were considered to have had an "informed" decision to be sent to the ED if they were seen in person or spoke with a doctor by phone. Pearson's Chi-Square testing was used to determine associations between the type of outpatient provider contact and ED disposition. A prospective cross-sectional study of adult ED patients who presented to a single tertiary care referral Level I trauma center with 115,000 annual ED visits. Consenting patients in the ED were surveyed during regular intervals during days and evenings 7 days/week. The survey was pilot-tested and validated in the target population. Patients were asked whether they had attempted to contact an outside provider prior to their ED visit, and if successful, what type of provider they contacted, whether it was in person or by phone, and what instructions they received. Responses were then matched with ED disposition data. Those patients physically seen in an office or medical aid unit, or told by phone to go the ED were considered "sent" to the ED. They were compared with the group that "self-triaged." Patients were considered to have had an "informed" decision to be sent to the ED if they were seen in person or spoke with a doctor by phone. Pearson's Chi-Square testing was used to determine associations between the type of outpatient provider contact and ED disposition. ResultsThere was no difference in admission rate between those who attempted to contact their doctors (98/207, 47%) and those who did not (122/151, 48%). There was no difference between those sent to the ED (86/166, 51%) and those who self-triaged (134/292, 46%). However, amongst those who were sent to the ED after phone contact, there was a higher admission rate in those who had spoken to a doctor (25/38, 66%) as compared to those who had spoken to a non-physician such as a receptionist, nurse, or midlevel provider (23/61, 38%), p<0.01. Patients who received an informed decision to be sent to the ED were more likely to be admitted (59/99, 60%) than others who were sent to the ED (23/61, 38%), p<0.01. There was no difference in admission rate between those who attempted to contact their doctors (98/207, 47%) and those who did not (122/151, 48%). There was no difference between those sent to the ED (86/166, 51%) and those who self-triaged (134/292, 46%). However, amongst those who were sent to the ED after phone contact, there was a higher admission rate in those who had spoken to a doctor (25/38, 66%) as compared to those who had spoken to a non-physician such as a receptionist, nurse, or midlevel provider (23/61, 38%), p<0.01. Patients who received an informed decision to be sent to the ED were more likely to be admitted (59/99, 60%) than others who were sent to the ED (23/61, 38%), p<0.01. ConclusionsPatients who seek outpatient care prior to their ED visit are admitted to the hospital just as often as those who self-triage. This supports recent studies demonstrating the inability of patients to determine the appropriateness of an ED visit. Some patients may be sent to an ED for testing or interventions not otherwise available as an outpatient. But, if admission rate is an indicator of appropriateness of ED visits, then a more informed decision to send the patient to the ED may reduce avoidable ED utilization. Patients who are physically seen by a provider, or at least speak to a physician by phone, are more likely to be admitted, compared with those who are directed to the ED by a non-physician over the phone. This suggests that improving patients' access to a physician, at least by phone, or expanding acute unscheduled care options for both assessments and advanced testing, may reduce ED visits that do not require hospital admission. Patients who seek outpatient care prior to their ED visit are admitted to the hospital just as often as those who self-triage. This supports recent studies demonstrating the inability of patients to determine the appropriateness of an ED visit. Some patients may be sent to an ED for testing or interventions not otherwise available as an outpatient. But, if admission rate is an indicator of appropriateness of ED visits, then a more informed decision to send the patient to the ED may reduce avoidable ED utilization. Patients who are physically seen by a provider, or at least speak to a physician by phone, are more likely to be admitted, compared with those who are directed to the ED by a non-physician over the phone. This suggests that improving patients' access to a physician, at least by phone, or expanding acute unscheduled care options for both assessments and advanced testing, may reduce ED visits that do not require hospital admission.

Study Objectives: The early recognition of time-sensitive, high mortality and destination-specific medical conditions is a key focus of emergency medical services systems. Whereas much attention has been focused on early and accurate out-of-hospital identification of ST-segment elevation MI, severely injured trauma patients and patients with acute stroke, there has been minimal attention directed to patients with sepsis. Mortality from severe sepsis and septic shock has been reported as high as 50%, with over half of potentially septic patients being transported by emergency medical services. Mounting evidence shows that early recognition and intervention in sepsis reduces morbidity and mortality. The goals of this study are: 1) determine the positive predictive value of an emergency medical services sepsis protocol utilizing point-of-care venous lactate to predict a final hospital diagnosis of severe infection or sepsis in out-of-hospital patients. 2) report differences in time to antibiotics and acuity in these patients utilizing a 2-tiered early notification and system response. 3) evaluate the correlation between out-of-hospital and hospital venous lactate level. Methods: Analysis of a prospective cohort of consecutive out-of-hospital patients treated under an emergency medical services sepsis protocol between July 2011 and March 2012 and transported to a large urban/suburban 2-hospital health system. Criteria for inclusion in the sepsis protocol were: 1) paramedic identification of patients at risk for sepsis using presence of 2 or more systemic inflammatory response syndrome (SIRS) criteria (continuous pulse rate>90, RR>20, T>38 or <36) plus clinical suspicion for infection, 2) out-of-hospital venous lactate measurement and 3) hospital notification of an “emergency medical services sepsis alert” (2 or more SIRS criteria, suspected infection, lactate ≥ 4 mmol/L) or “emergency medical services sepsis advisory” (2 or more SIRS criteria, suspected infection, lactate 2.5–3.9 mmol/L) prior to or at hospital arrival. Hospital response was tiered based on “alert” or “advisory” notification, with sepsis alert patients receiving immediate bedding and physician evaluation. Results: A total of 219 patients met the threshold lactate level for inclusion. 36 patients did not meet SIRS criteria or have documented hospital notification and were excluded from analysis, leaving 86 “emergency medical services sepsis alert” and 97 “emergency medical services sepsis advisory” patients. 76.7% of sepsis alert patients (n=66) and 74.2% of sepsis advisory patients (n=72) had a final hospital diagnosis of severe infection or sepsis. In these patients, median time of arrival to broad-spectrum antibiotics was 59 min (IQR=42–91) in sepsis alert patients and 81 min (IQR=49.5–127.3) in sepsis advisory patients. ICU admission occurred in 50% and 23% of sepsis alert and advisory, respectively. Median out-of-hospital and ED lactate levels were comparable and these levels demonstrated strong correlation by a Pearson correlation coefficient (r)=0.8. Conclusion: A out-of-hospital sepsis protocol accurately predicts severe infection and sepsis. The potential time savings from early out-of-hospital recognition, notification and intervention in patients with sepsis is likely to have a positive impact on morbidity and mortality. Further investigation of out-of-hospital sepsis interventions, including clinical outcomes and identifying optimal lactate cut-off values are warranted to better evaluate effectiveness and optimize resource utilization.

BackgroundIt has been shown that crowding in emergency departments (EDs) contributes to increased ED and hospital length of stay (LOS), increased patient mortality, lost hospital revenue, and delays in treatment such as timely administration of antibiotics for pneumonia or medication for pain.Study ObjectivesTo determine if an association exists between specific indicators of ED crowding and the number of cases originating each day for the performance improvement (PI) review process in a single hospital system.MethodsA retrospective analytic cohort study of PI cases generated in one hospital system (comprised of 2 hospitals- one urban, one suburban; combined ED volume > 150,000) was conducted. For each site the number of PI cases originated per calendar day of 2008 was collected, as well as the following daily data: number of arrivals to the ED, number of patients who left without treatment (LWOT), total boarding hours, average boarding time, number of admissions through the ED, average ED LOS, average ED occupancy, hospital occupancy, and average emergency severity index (ESI) score. An institution-specific triage crowding score based on the median number of patients waiting at triage for 3 separate time periods (0700-1500, 1500-2300, and 2300-0700) was also calculated for each study day. Spearman's correlation coefficient was utilized to determine statistical dependence between the number of PI cases originated per day and the above crowding factors at each site. A p-value of <0.01 was considered statistically significant.ResultsDuring 2008 there were 503 PI cases originated out of 106,035 patient visits at the suburban site (0.47%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ 0.131, p 0.012), number of LWOTs (ñ 0.094, p 0.074), total boarding hours (ñ 0.094, p 0.074), average boarding time (ñ 0.076, p 0.145), average ED LOS (ñ 0.121, p 0.021), average ED occupancy (ñ 0.025, p 0.633), average hospital occupancy (ñ -0.091, p 0.081), triage crowding score (ñ 0.079, p 0.133), or average ESI score (ñ 0.015, p 0.775). There was a weak, though statistically significant correlation between number of hospital admissions from the ED and number of PI cases generated (ñ 0.144, p 0.006). At the urban site there were 50,844 people seen and 113 PI cases generated (0.22%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ -0.106 , p 0.042), number of LWOTs (ñ -0.029, p 0.575), total boarding hours (ñ-0.075, p 0.151), average boarding time (ñ-0.050, p 0.344), total number of hospital admissions from the ED (ñ-0.104, p 0.046), average ED LOS (ñ -0.010, p 0.844), average ED occupancy (ñ -0.090, p 0.085), average hospital occupancy (ñ 0.036, p 0.497), triage crowding score (ñ-0.038, p 0.472), or average ESI score (ñ 0.025, p 0.636). Additional analysis showed that the above crowding factors did not significantly differ at either site between days with and without PI cases generated.ConclusionsIncreased crowding at these two EDs did not appear to correlate with increased generation of PI cases. Additional multi-center studies would be useful to determine if this is a site-specific conclusion. If so, it would be beneficial to examine the procedures in place which allow some sites to effectively compensate for crowding. BackgroundIt has been shown that crowding in emergency departments (EDs) contributes to increased ED and hospital length of stay (LOS), increased patient mortality, lost hospital revenue, and delays in treatment such as timely administration of antibiotics for pneumonia or medication for pain. It has been shown that crowding in emergency departments (EDs) contributes to increased ED and hospital length of stay (LOS), increased patient mortality, lost hospital revenue, and delays in treatment such as timely administration of antibiotics for pneumonia or medication for pain. Study ObjectivesTo determine if an association exists between specific indicators of ED crowding and the number of cases originating each day for the performance improvement (PI) review process in a single hospital system. To determine if an association exists between specific indicators of ED crowding and the number of cases originating each day for the performance improvement (PI) review process in a single hospital system. MethodsA retrospective analytic cohort study of PI cases generated in one hospital system (comprised of 2 hospitals- one urban, one suburban; combined ED volume > 150,000) was conducted. For each site the number of PI cases originated per calendar day of 2008 was collected, as well as the following daily data: number of arrivals to the ED, number of patients who left without treatment (LWOT), total boarding hours, average boarding time, number of admissions through the ED, average ED LOS, average ED occupancy, hospital occupancy, and average emergency severity index (ESI) score. An institution-specific triage crowding score based on the median number of patients waiting at triage for 3 separate time periods (0700-1500, 1500-2300, and 2300-0700) was also calculated for each study day. Spearman's correlation coefficient was utilized to determine statistical dependence between the number of PI cases originated per day and the above crowding factors at each site. A p-value of <0.01 was considered statistically significant. A retrospective analytic cohort study of PI cases generated in one hospital system (comprised of 2 hospitals- one urban, one suburban; combined ED volume > 150,000) was conducted. For each site the number of PI cases originated per calendar day of 2008 was collected, as well as the following daily data: number of arrivals to the ED, number of patients who left without treatment (LWOT), total boarding hours, average boarding time, number of admissions through the ED, average ED LOS, average ED occupancy, hospital occupancy, and average emergency severity index (ESI) score. An institution-specific triage crowding score based on the median number of patients waiting at triage for 3 separate time periods (0700-1500, 1500-2300, and 2300-0700) was also calculated for each study day. Spearman's correlation coefficient was utilized to determine statistical dependence between the number of PI cases originated per day and the above crowding factors at each site. A p-value of <0.01 was considered statistically significant. ResultsDuring 2008 there were 503 PI cases originated out of 106,035 patient visits at the suburban site (0.47%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ 0.131, p 0.012), number of LWOTs (ñ 0.094, p 0.074), total boarding hours (ñ 0.094, p 0.074), average boarding time (ñ 0.076, p 0.145), average ED LOS (ñ 0.121, p 0.021), average ED occupancy (ñ 0.025, p 0.633), average hospital occupancy (ñ -0.091, p 0.081), triage crowding score (ñ 0.079, p 0.133), or average ESI score (ñ 0.015, p 0.775). There was a weak, though statistically significant correlation between number of hospital admissions from the ED and number of PI cases generated (ñ 0.144, p 0.006). At the urban site there were 50,844 people seen and 113 PI cases generated (0.22%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ -0.106 , p 0.042), number of LWOTs (ñ -0.029, p 0.575), total boarding hours (ñ-0.075, p 0.151), average boarding time (ñ-0.050, p 0.344), total number of hospital admissions from the ED (ñ-0.104, p 0.046), average ED LOS (ñ -0.010, p 0.844), average ED occupancy (ñ -0.090, p 0.085), average hospital occupancy (ñ 0.036, p 0.497), triage crowding score (ñ-0.038, p 0.472), or average ESI score (ñ 0.025, p 0.636). Additional analysis showed that the above crowding factors did not significantly differ at either site between days with and without PI cases generated. During 2008 there were 503 PI cases originated out of 106,035 patient visits at the suburban site (0.47%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ 0.131, p 0.012), number of LWOTs (ñ 0.094, p 0.074), total boarding hours (ñ 0.094, p 0.074), average boarding time (ñ 0.076, p 0.145), average ED LOS (ñ 0.121, p 0.021), average ED occupancy (ñ 0.025, p 0.633), average hospital occupancy (ñ -0.091, p 0.081), triage crowding score (ñ 0.079, p 0.133), or average ESI score (ñ 0.015, p 0.775). There was a weak, though statistically significant correlation between number of hospital admissions from the ED and number of PI cases generated (ñ 0.144, p 0.006). At the urban site there were 50,844 people seen and 113 PI cases generated (0.22%). There was no statistically significant correlation between the number of PI cases generated each day and total arrivals to the ED (ñ -0.106 , p 0.042), number of LWOTs (ñ -0.029, p 0.575), total boarding hours (ñ-0.075, p 0.151), average boarding time (ñ-0.050, p 0.344), total number of hospital admissions from the ED (ñ-0.104, p 0.046), average ED LOS (ñ -0.010, p 0.844), average ED occupancy (ñ -0.090, p 0.085), average hospital occupancy (ñ 0.036, p 0.497), triage crowding score (ñ-0.038, p 0.472), or average ESI score (ñ 0.025, p 0.636). Additional analysis showed that the above crowding factors did not significantly differ at either site between days with and without PI cases generated. ConclusionsIncreased crowding at these two EDs did not appear to correlate with increased generation of PI cases. Additional multi-center studies would be useful to determine if this is a site-specific conclusion. If so, it would be beneficial to examine the procedures in place which allow some sites to effectively compensate for crowding. Increased crowding at these two EDs did not appear to correlate with increased generation of PI cases. Additional multi-center studies would be useful to determine if this is a site-specific conclusion. If so, it would be beneficial to examine the procedures in place which allow some sites to effectively compensate for crowding.

Study ObjectivesCrowding of emergency departments has required physicians to evaluate and treat patients in overflow areas, including hallways, which can be inefficient and frustrating. Novel approaches have included a centralized team of providers with mobile patients versus a traditional roving team that evaluates the patient at the hallway site. This approach has been shown to have no immediate degradation in productivity with implementation, but is associated with increased provider satisfaction. Follow-up analysis of the original group at one year after implementation was performed to evaluate possible increased productivity over time.MethodsEmergency physicians working in an academic setting with over 100,000 annual visits were evaluated. Physician productivity was evaluated by Patients/hour (Pts/h), Relative Value Units/hour (RVU/h), and Charges/hour (Chg/h) before, after the model change and one year following. Physicians were also administered a survey tool with a ten-point visual analog scale that was validated by qualitative methodology to evaluate level of satisfaction and perceived productivity in the different practice models.ResultsPhysician productivity in a hallway model (Pts/h 1.64+0.37, RVU/h 4.69+0.98, Chg/h $481+98) versus centralized model (Pts/h 1.77+0.49, RVU/h 4.85+1.16, Chg/h $508+122) showed no statistically significance difference, (p>0.05) as previously reported. One year later the centralized model (Pts/h 1.78+0.39, RVU/h 4.77 +1.16, Chg/h $506+116) showed no statistical difference to the hallway model or the initial centralized model (p>0.05). Physicians rated their perceived productivity higher in the centralized model 3.55+0.83 versus 1.43+0.93, (p-value of 0.001). The physician satisfaction was also higher in the centralized model 3.61+0.90 versus 1.12 +1.03, (p-value 0.001).ConclusionWith an implementation of a new workflow pattern utilizing the same number of health care providers and resources leads to higher satisfaction and perceived productivity. While initial implementation shows no degradation in productivity there is a lack of increased productivity over time. While changes in workflow patterns can improve perceived productivity and satisfaction they may only have limited effects on actual productivity as measured by standard parameters. Alterations to workflow management need to be assessed by objective criteria in order to assess for true changes in productivity and workflow impact. Study ObjectivesCrowding of emergency departments has required physicians to evaluate and treat patients in overflow areas, including hallways, which can be inefficient and frustrating. Novel approaches have included a centralized team of providers with mobile patients versus a traditional roving team that evaluates the patient at the hallway site. This approach has been shown to have no immediate degradation in productivity with implementation, but is associated with increased provider satisfaction. Follow-up analysis of the original group at one year after implementation was performed to evaluate possible increased productivity over time. Crowding of emergency departments has required physicians to evaluate and treat patients in overflow areas, including hallways, which can be inefficient and frustrating. Novel approaches have included a centralized team of providers with mobile patients versus a traditional roving team that evaluates the patient at the hallway site. This approach has been shown to have no immediate degradation in productivity with implementation, but is associated with increased provider satisfaction. Follow-up analysis of the original group at one year after implementation was performed to evaluate possible increased productivity over time. MethodsEmergency physicians working in an academic setting with over 100,000 annual visits were evaluated. Physician productivity was evaluated by Patients/hour (Pts/h), Relative Value Units/hour (RVU/h), and Charges/hour (Chg/h) before, after the model change and one year following. Physicians were also administered a survey tool with a ten-point visual analog scale that was validated by qualitative methodology to evaluate level of satisfaction and perceived productivity in the different practice models. Emergency physicians working in an academic setting with over 100,000 annual visits were evaluated. Physician productivity was evaluated by Patients/hour (Pts/h), Relative Value Units/hour (RVU/h), and Charges/hour (Chg/h) before, after the model change and one year following. Physicians were also administered a survey tool with a ten-point visual analog scale that was validated by qualitative methodology to evaluate level of satisfaction and perceived productivity in the different practice models. ResultsPhysician productivity in a hallway model (Pts/h 1.64+0.37, RVU/h 4.69+0.98, Chg/h $481+98) versus centralized model (Pts/h 1.77+0.49, RVU/h 4.85+1.16, Chg/h $508+122) showed no statistically significance difference, (p>0.05) as previously reported. One year later the centralized model (Pts/h 1.78+0.39, RVU/h 4.77 +1.16, Chg/h $506+116) showed no statistical difference to the hallway model or the initial centralized model (p>0.05). Physicians rated their perceived productivity higher in the centralized model 3.55+0.83 versus 1.43+0.93, (p-value of 0.001). The physician satisfaction was also higher in the centralized model 3.61+0.90 versus 1.12 +1.03, (p-value 0.001). Physician productivity in a hallway model (Pts/h 1.64+0.37, RVU/h 4.69+0.98, Chg/h $481+98) versus centralized model (Pts/h 1.77+0.49, RVU/h 4.85+1.16, Chg/h $508+122) showed no statistically significance difference, (p>0.05) as previously reported. One year later the centralized model (Pts/h 1.78+0.39, RVU/h 4.77 +1.16, Chg/h $506+116) showed no statistical difference to the hallway model or the initial centralized model (p>0.05). Physicians rated their perceived productivity higher in the centralized model 3.55+0.83 versus 1.43+0.93, (p-value of 0.001). The physician satisfaction was also higher in the centralized model 3.61+0.90 versus 1.12 +1.03, (p-value 0.001). ConclusionWith an implementation of a new workflow pattern utilizing the same number of health care providers and resources leads to higher satisfaction and perceived productivity. While initial implementation shows no degradation in productivity there is a lack of increased productivity over time. While changes in workflow patterns can improve perceived productivity and satisfaction they may only have limited effects on actual productivity as measured by standard parameters. Alterations to workflow management need to be assessed by objective criteria in order to assess for true changes in productivity and workflow impact. With an implementation of a new workflow pattern utilizing the same number of health care providers and resources leads to higher satisfaction and perceived productivity. While initial implementation shows no degradation in productivity there is a lack of increased productivity over time. While changes in workflow patterns can improve perceived productivity and satisfaction they may only have limited effects on actual productivity as measured by standard parameters. Alterations to workflow management need to be assessed by objective criteria in order to assess for true changes in productivity and workflow impact.