- Email: [email protected]
77 FastER Care in the Emergency Department Leads to Improvement in Emergency Department Throughput Metrics as well as Improved Patient Experience
Research Forum Abstracts Conclusions: A modified TIMI score of zero or a HEART score of 2, incorporating a single hs-cTnT level, will identify patients with low risk of 30-day MACE for early discharge within 2 hours of ED arrival.
74
Mean HEART Scores for Admitted Chest Pain Patients are Higher in More Experienced Providers
Dubin J, Kiechle E, Wilson M, Timbol C, Bhat R, Milzman D/WHC, washington, DC
Study Objectives: The History, EKG, Age, Risk factors and Troponin (HEART) score has been validated as a predictor of acute coronary syndrome (ACS) and major adverse cardiac event (MACE) in emergency department (ED) patients complaining of chest pain. Our objective was to determine the extent of physician variation in the HEART score of admitted patients, and to determine whether duration of practice and residency training institution affected the HEART score or MACE rate in admitted patients. Methods: This was a retrospective chart review at a busy mid-Atlantic ED with a yearly census of 90,000 patients. Data, including HEART score, outcome of admission and 30-day MACE, were abstracted by one of four authors, with EKGs being graded by consensus between three physicians. We used analysis of variance to determine the difference in mean HEART scores between providers, Fisher’s exact test to determine difference in MACE by duration of training, and logistic regression to determine predictors of low risk admission (HEART Score 3). Results: We reviewed 585 chest pain admissions, including at least 30 admissions for each of 19 full time board certified/board elligible physicians working backwards from December 2014. The average mean HEART score was 4.41 (SD 0.43), with a physician range of 3.66 to 5.10. Taken individually there was no difference in physician mean scores (p¼0.070), but physicians with 10-15 years of experience had significantly higher mean scores than those with 0-5 years of experience (mean HEART Score 4.65 vs 3.93, p¼0.012). Those with 10-15 years of experience also had significantly higher proportion of MACE (15.3%, p¼0.002). Less than 5 years of postresidency training was a predictor of low risk admission in logistic regression (p¼0.028). There was no difference in mean HEART Score between training institutions represented by at least 2 full time physicans (p¼0.058). Conclusion: In our analysis, more experienced providers admitted higher risk patients and were more likely to admit patients who would experience a MACE. More research is needed to determine whether this decisionmaking can be used to increase sensitivity for admitting patients at high risk for MACE, and for decreasing admissions for ACS workup in low risk patients.
75 76
Withdrawn
Effect of a Sickle Cell Vaso-occlusive Crisis Observation Unit Pathway on the Admission Rate for Frequent Emergency Department Users
Loeffler P, Sturgis L, Muelller T, Gibson R, Lyon M/Augusta University, Augusta, GA
Study Objectives: Vaso-occlusive events (VOE) is the most common complication of sickle cell disease (SCD) and the most frequent reason patients seek care in the emergency department (ED). Individualized protocol-based pathways have been used in ED observation units (EDOUs) to standardize and improve the quality of care for patients with uncomplicated VOE due to SCD. Frequency of visits to the ED is often used as a proxy for disease severity. A previous study of EDOU utilization by patients with uncomplicated VOE found different admission rates across utilization groups. The objective of this study was to evaluate the relationship between the number of EDOU visit and rate of admission. Methods: This was a retrospective chart review of adult patients with uncomplicated VOE (no fever, hypoxia, acute chest, or evidence of sepsis) due to SCD who were treated at an urban tertiary care hospital between 9/2013 and 5/2015. All patients presenting to the ED with uncomplicated VOE due to SCD are treated in the EDOU using a standardized VOC pathway that includes an individualized opioid treatment plan and patient controlled analgesia (PCA) pump. Patients are treated on
S32 Annals of Emergency Medicine
this pathway for up to 24 hours. If their pain is not controlled within 24 hours or if a complication develops they are admitted to the hospital. Patients were classified as high users (>4 visits/year), moderate users (1-3 visits/year) or low users (<1 visit/year) based on average annual EDOU utilization during the study period. Rates of admission for each user group were compared using an unpaired, one-tailed, Student’s t-test. Results: A total of 727 visits for 154 patients were included in the analysis. High-users (44) had a total of 539 visits and an average patient admission rate of 22% (n¼118). Moderate-users (49) had a total of 108 visits and an admission rate of 31% (n¼33). Low-users (61) had a total of 80 visits and an admission rate of 36% (n¼29). The difference in the admission rate between high-users and low-users was significant (p<0.01) as was the difference between the average number of visits of moderate users and high-users (p¼0.04). The difference in the average admission rate of moderate users and low users was not significant (p¼0.14). Conclusion: There are many factors that may influence the rate of ED/EDOU utilization by patients with SCD for uncomplicated VOE. The results of this study show a strong negative correlation between EDOU utilization and hospital admission rates. The results of this study support the hypothesis that an EDOU-based standardized pathway has a greater effect on reducing admission rates for patients who use the EDOU frequently (>4 visits/year) than those who use the EDOU less frequently (<4 visits/year) for pain control. A possible explanation for this is that in SCD VOE, ED utilization for pain may not be an indicator of disease severity. Future studies will investigate other factors that influence ED utilization by patients with SCD for VOC.
77
FastER Care in the Emergency Department Leads to Improvement in Emergency Department Throughput Metrics as well as Improved Patient Experience
Robinson M, Sampson A, Ticgelaor J/University of Missouri-Columbia, Columbia, MO
Study Objective: The Hospital Assessment of Healthcare Providers and Systems (HCAHPS) public reporting initiative mandated by the Centers for Medicare and Medicaid Services (CMS) requires reporting of multiple quality measures. These include emergency department (ED) throughput measures as well as ratings of the patient experience in the ED. These ratings are shared with the public, and not only affect the hospital’s reputation, but are also used in calculating value-based purchasing payments. In an effort to improve throughput metrics, split-flow processes such as a Fast Tracks are used to rapidly treat lower acuity patients in a more expeditious manner. Current literature has shown Fast Tracks improve patient flow within the ED. However, there are few studies examining the effect of a Fast Track on patient experience alone, or in combination with throughput metrics. We aimed to evaluate the impact of creating a Fast Track on both patient experience and ED throughput metrics. Methods: This was an IRB-approved retrospective review of ED patients presenting to the University of Missouri-Columbia Health Center from December 1, 2015 through April 1, 2016. Our Fast Track (FastER Care) consisted of a single treatment room with an internal waiting room. Evaluations occurred in the private room, and patients waited for test results in the internal waiting room. Patients triaged as a level 4 or 5 using the Emergency Severity Index (ESI) with focused complaints were eligible for care in this area. Limited laboratory and plain radiographs were available. Patients who were initially triaged to FastER Care were immediately moved to the main ED if they needed a higher level of care. The area was staffed daily for 11 hours by a physician’s assistant and ED nurse. No staffing increases were used to implement the Fast Track, and no other ED initiatives were started during the study period. Patient satisfaction was measured using the National Research Corporation Picker Survey. Results: During the study period, a total of 16,511 ED patients checked into the ED. 12,358 patients checked in during the hours of operation of FastER. A total of 2201 patients were seen in the Fast Track, representing 13.3% of the total ED patients, and 17.8% of patients presenting during hours of operation. A simple linear regression was created to predict throughput measures based on monthly ED census, and compared to actual metrics for the four months following implementation as follows: The average weekly patient satisfaction rate for the 5 months prior to implementation of FastER care was 32.0%. During the study period, average weekly satisfaction increased to 61.8%.
Volume 68, no. 4s : October 2016
Research Forum Abstracts
Dec Predicted
Dec Actual
Jan Predicted
Jan Actual
Feb Predicted
Feb Actual
Mar Predicted
Mar Actual
1.22
0.62
1.58
0.99
1.64
1.3
2.16
1.82
Door To Doc
16
15
17
15
18
17
20
19
LOS Dicharged
160
156
167
159
169
167
178
175
LWBS
Conclusion: CMS has many quality measures including ED length of stay and patient satisfaction. The implementation of a Fast Track simultaneously improved both of these metrics within our emergency department. This study suggests that the use of a Fast Track effectively improves ED workflow without damaging the patients’ experience. Expanding the area and improving throughput time may add even greater benefit.
78
What Happens to Emergency Patients Who Leave Without Being Seen by a Physician?
Vilke GM, Castillo EM, Brennan JJ, Killeen JP/University of California, San Diego, San Diego, CA
Study Objective: With increasing capacity issues in the emergency department, patients are experiencing longer wait times and often because of this, many patients leave without being seen (LWBS) by a physician. Methods: Retrospective review of an electronic medical record databases to assess for all patients who checked in to two teaching hospitals (one urban and one community based) with a combined annual ED census of approximately 75,000 annual visits. The database was searched for all patients who checked into the ED triage, but then left the ED before being seen by the physician. The study period included all visits between July 1, 2015 and February 29, 2016. The database was then queried for whether the patient returned to either of the two emergency departments. 24, 48, 72 and 168-hour return visits were calculated as the number of minutes between the ED departure time of the index visit and ED arrival time of return visit. Descriptive statistics were utilized. Results: During the seven-month study period, there were a total of 50766 ED visits, of which 1847 (3.6%) patients LWBS. 308 (16.7%) of these patients returned within 24 hours to one of the EDs for evaluation. This number went up to 374 (20.2%) at 48 hours, 411 (22.3%) at 72 hours and 501 (27.1%) returned within 7 days. There was an increased percentage of patients who LWBS from the urban ED (4.9%) compared with the community hospital (1.5%). Additionally the percentage of those who returned in 1 and 7 days was greater at the urban hospital (17.6% and 28.9%) compared with the community hospital (11.5% and 22.0%). Conclusions: A significant percentage of patients LWBS from the emergency department, 3.6% in our study. However, more than a quarter of them return to the ED within a week of leaving. Future research should seek to assess this population and their outcomes - are they the mild and truly unsick, or are they patients who are so unwell that they cannot tolerate long waits in a crowded waiting room?
79
F ¼ 0.0001 R2¼ .84 P<0.05 F¼0.0005 R2¼.84 P<0.05 F¼ 0.0005 R2¼.81 P<0.005
Methods: This study was a retrospective database review of a database of patients admitted to the EDOU on a SCP for VOC. Data were analyzed for a 21-month period (December, 2011 - August, 2013) in which patient care was managed by emergency physicianss and a 21-month period (September, 2013 -May, 2015) in which patient care was managed by hospitalists physicians. During these two time periods, there were no changes in nursing management or staffing and no changes in the SCP. Hospitalists physician management included Advance Practice Providers (APPs) while the emergency physician management did not. 773 patient encounters were included from the emergency physician period and 727 from the hospitalists physican period. Data extracted included: length of stay (LOS), disposition, three and 30-day return, and readmission rates. Only uncomplicated VOC visits due to SCD presenting to the EDOU were included in this study. Results: The average EDOU length of stay (LOS) for a patient with SCD during the emergency physician management period was 17 hours and 54 minutes; during the hospitalists physician management period, the average LOS was 18 hours and 23 minutes. Data were analyzed by patient disposition, and three and 30-day return. The admission rate during the emergency physician management period was statistically different than the admission rate during the hospitalists physican management period (16% versus 24.8% p<0.0001). There was no statistically significant difference in the three-day return rates and the 30-day return rates. Of the patients initially admitted to the hospital, the 30-day readmission rate under the emergency physician management period was 8.9% while it was 15.6% for the hospitalists physician management period (p<0.001) (not included in the figure). The EDOU LOS was not statistically different between groups (p¼.1853). Conclusion: While there are many factors that affect the rate of admission and ED utilization for uncomplicated VOC by patients with SCD. This research shows that there were both statistically significant and clinically significant differences in the hospital admission rates of patients with VOC during two time periods where care was managed by two different groups (emergency physicians and hospitalists physicians). Future research will examine the reasons for these differences.
Management Group Admittance Rates from an Emergency Department Observation Unit for Patients With Sickle Cell Disease
Mueller T, Sturgis L, Loeffler P, Kuchinski AM, Robert G, Lyon M/Augusta University, Augusta, GA
Study Objectives: Emergency department observation units (EDOUs) and sickle cell pathways (SCPs) allow for protocol-based rapid initiation of analgesic treatment for patients with sickle cell disease (SCD) during a vaso-occlusive crisis (VOC). There are many variations in the staffing and management of EDOUs. Our primary objective in this study was to compare hospital admission rates of patients treated in an EDOU for VOC due to SCD when managed by two different groups: Emergency physicians and hospitalists physician. In addition, we compared the three and 30-day return rates of patients when managed by these two different groups as a measure of premature discharge from the SCP.
Volume 68, no. 4s : October 2016
Annals of Emergency Medicine S33
74
Mean HEART Scores for Admitted Chest Pain Patients are Higher in More Experienced Providers
Dubin J, Kiechle E, Wilson M, Timbol C, Bhat R, Milzman D/WHC, washington, DC
Study Objectives: The History, EKG, Age, Risk factors and Troponin (HEART) score has been validated as a predictor of acute coronary syndrome (ACS) and major adverse cardiac event (MACE) in emergency department (ED) patients complaining of chest pain. Our objective was to determine the extent of physician variation in the HEART score of admitted patients, and to determine whether duration of practice and residency training institution affected the HEART score or MACE rate in admitted patients. Methods: This was a retrospective chart review at a busy mid-Atlantic ED with a yearly census of 90,000 patients. Data, including HEART score, outcome of admission and 30-day MACE, were abstracted by one of four authors, with EKGs being graded by consensus between three physicians. We used analysis of variance to determine the difference in mean HEART scores between providers, Fisher’s exact test to determine difference in MACE by duration of training, and logistic regression to determine predictors of low risk admission (HEART Score 3). Results: We reviewed 585 chest pain admissions, including at least 30 admissions for each of 19 full time board certified/board elligible physicians working backwards from December 2014. The average mean HEART score was 4.41 (SD 0.43), with a physician range of 3.66 to 5.10. Taken individually there was no difference in physician mean scores (p¼0.070), but physicians with 10-15 years of experience had significantly higher mean scores than those with 0-5 years of experience (mean HEART Score 4.65 vs 3.93, p¼0.012). Those with 10-15 years of experience also had significantly higher proportion of MACE (15.3%, p¼0.002). Less than 5 years of postresidency training was a predictor of low risk admission in logistic regression (p¼0.028). There was no difference in mean HEART Score between training institutions represented by at least 2 full time physicans (p¼0.058). Conclusion: In our analysis, more experienced providers admitted higher risk patients and were more likely to admit patients who would experience a MACE. More research is needed to determine whether this decisionmaking can be used to increase sensitivity for admitting patients at high risk for MACE, and for decreasing admissions for ACS workup in low risk patients.
75 76
Withdrawn
Effect of a Sickle Cell Vaso-occlusive Crisis Observation Unit Pathway on the Admission Rate for Frequent Emergency Department Users
Loeffler P, Sturgis L, Muelller T, Gibson R, Lyon M/Augusta University, Augusta, GA
Study Objectives: Vaso-occlusive events (VOE) is the most common complication of sickle cell disease (SCD) and the most frequent reason patients seek care in the emergency department (ED). Individualized protocol-based pathways have been used in ED observation units (EDOUs) to standardize and improve the quality of care for patients with uncomplicated VOE due to SCD. Frequency of visits to the ED is often used as a proxy for disease severity. A previous study of EDOU utilization by patients with uncomplicated VOE found different admission rates across utilization groups. The objective of this study was to evaluate the relationship between the number of EDOU visit and rate of admission. Methods: This was a retrospective chart review of adult patients with uncomplicated VOE (no fever, hypoxia, acute chest, or evidence of sepsis) due to SCD who were treated at an urban tertiary care hospital between 9/2013 and 5/2015. All patients presenting to the ED with uncomplicated VOE due to SCD are treated in the EDOU using a standardized VOC pathway that includes an individualized opioid treatment plan and patient controlled analgesia (PCA) pump. Patients are treated on
S32 Annals of Emergency Medicine
this pathway for up to 24 hours. If their pain is not controlled within 24 hours or if a complication develops they are admitted to the hospital. Patients were classified as high users (>4 visits/year), moderate users (1-3 visits/year) or low users (<1 visit/year) based on average annual EDOU utilization during the study period. Rates of admission for each user group were compared using an unpaired, one-tailed, Student’s t-test. Results: A total of 727 visits for 154 patients were included in the analysis. High-users (44) had a total of 539 visits and an average patient admission rate of 22% (n¼118). Moderate-users (49) had a total of 108 visits and an admission rate of 31% (n¼33). Low-users (61) had a total of 80 visits and an admission rate of 36% (n¼29). The difference in the admission rate between high-users and low-users was significant (p<0.01) as was the difference between the average number of visits of moderate users and high-users (p¼0.04). The difference in the average admission rate of moderate users and low users was not significant (p¼0.14). Conclusion: There are many factors that may influence the rate of ED/EDOU utilization by patients with SCD for uncomplicated VOE. The results of this study show a strong negative correlation between EDOU utilization and hospital admission rates. The results of this study support the hypothesis that an EDOU-based standardized pathway has a greater effect on reducing admission rates for patients who use the EDOU frequently (>4 visits/year) than those who use the EDOU less frequently (<4 visits/year) for pain control. A possible explanation for this is that in SCD VOE, ED utilization for pain may not be an indicator of disease severity. Future studies will investigate other factors that influence ED utilization by patients with SCD for VOC.
77
FastER Care in the Emergency Department Leads to Improvement in Emergency Department Throughput Metrics as well as Improved Patient Experience
Robinson M, Sampson A, Ticgelaor J/University of Missouri-Columbia, Columbia, MO
Study Objective: The Hospital Assessment of Healthcare Providers and Systems (HCAHPS) public reporting initiative mandated by the Centers for Medicare and Medicaid Services (CMS) requires reporting of multiple quality measures. These include emergency department (ED) throughput measures as well as ratings of the patient experience in the ED. These ratings are shared with the public, and not only affect the hospital’s reputation, but are also used in calculating value-based purchasing payments. In an effort to improve throughput metrics, split-flow processes such as a Fast Tracks are used to rapidly treat lower acuity patients in a more expeditious manner. Current literature has shown Fast Tracks improve patient flow within the ED. However, there are few studies examining the effect of a Fast Track on patient experience alone, or in combination with throughput metrics. We aimed to evaluate the impact of creating a Fast Track on both patient experience and ED throughput metrics. Methods: This was an IRB-approved retrospective review of ED patients presenting to the University of Missouri-Columbia Health Center from December 1, 2015 through April 1, 2016. Our Fast Track (FastER Care) consisted of a single treatment room with an internal waiting room. Evaluations occurred in the private room, and patients waited for test results in the internal waiting room. Patients triaged as a level 4 or 5 using the Emergency Severity Index (ESI) with focused complaints were eligible for care in this area. Limited laboratory and plain radiographs were available. Patients who were initially triaged to FastER Care were immediately moved to the main ED if they needed a higher level of care. The area was staffed daily for 11 hours by a physician’s assistant and ED nurse. No staffing increases were used to implement the Fast Track, and no other ED initiatives were started during the study period. Patient satisfaction was measured using the National Research Corporation Picker Survey. Results: During the study period, a total of 16,511 ED patients checked into the ED. 12,358 patients checked in during the hours of operation of FastER. A total of 2201 patients were seen in the Fast Track, representing 13.3% of the total ED patients, and 17.8% of patients presenting during hours of operation. A simple linear regression was created to predict throughput measures based on monthly ED census, and compared to actual metrics for the four months following implementation as follows: The average weekly patient satisfaction rate for the 5 months prior to implementation of FastER care was 32.0%. During the study period, average weekly satisfaction increased to 61.8%.
Volume 68, no. 4s : October 2016
Research Forum Abstracts
Dec Predicted
Dec Actual
Jan Predicted
Jan Actual
Feb Predicted
Feb Actual
Mar Predicted
Mar Actual
1.22
0.62
1.58
0.99
1.64
1.3
2.16
1.82
Door To Doc
16
15
17
15
18
17
20
19
LOS Dicharged
160
156
167
159
169
167
178
175
LWBS
Conclusion: CMS has many quality measures including ED length of stay and patient satisfaction. The implementation of a Fast Track simultaneously improved both of these metrics within our emergency department. This study suggests that the use of a Fast Track effectively improves ED workflow without damaging the patients’ experience. Expanding the area and improving throughput time may add even greater benefit.
78
What Happens to Emergency Patients Who Leave Without Being Seen by a Physician?
Vilke GM, Castillo EM, Brennan JJ, Killeen JP/University of California, San Diego, San Diego, CA
Study Objective: With increasing capacity issues in the emergency department, patients are experiencing longer wait times and often because of this, many patients leave without being seen (LWBS) by a physician. Methods: Retrospective review of an electronic medical record databases to assess for all patients who checked in to two teaching hospitals (one urban and one community based) with a combined annual ED census of approximately 75,000 annual visits. The database was searched for all patients who checked into the ED triage, but then left the ED before being seen by the physician. The study period included all visits between July 1, 2015 and February 29, 2016. The database was then queried for whether the patient returned to either of the two emergency departments. 24, 48, 72 and 168-hour return visits were calculated as the number of minutes between the ED departure time of the index visit and ED arrival time of return visit. Descriptive statistics were utilized. Results: During the seven-month study period, there were a total of 50766 ED visits, of which 1847 (3.6%) patients LWBS. 308 (16.7%) of these patients returned within 24 hours to one of the EDs for evaluation. This number went up to 374 (20.2%) at 48 hours, 411 (22.3%) at 72 hours and 501 (27.1%) returned within 7 days. There was an increased percentage of patients who LWBS from the urban ED (4.9%) compared with the community hospital (1.5%). Additionally the percentage of those who returned in 1 and 7 days was greater at the urban hospital (17.6% and 28.9%) compared with the community hospital (11.5% and 22.0%). Conclusions: A significant percentage of patients LWBS from the emergency department, 3.6% in our study. However, more than a quarter of them return to the ED within a week of leaving. Future research should seek to assess this population and their outcomes - are they the mild and truly unsick, or are they patients who are so unwell that they cannot tolerate long waits in a crowded waiting room?
79
F ¼ 0.0001 R2¼ .84 P<0.05 F¼0.0005 R2¼.84 P<0.05 F¼ 0.0005 R2¼.81 P<0.005
Methods: This study was a retrospective database review of a database of patients admitted to the EDOU on a SCP for VOC. Data were analyzed for a 21-month period (December, 2011 - August, 2013) in which patient care was managed by emergency physicianss and a 21-month period (September, 2013 -May, 2015) in which patient care was managed by hospitalists physicians. During these two time periods, there were no changes in nursing management or staffing and no changes in the SCP. Hospitalists physician management included Advance Practice Providers (APPs) while the emergency physician management did not. 773 patient encounters were included from the emergency physician period and 727 from the hospitalists physican period. Data extracted included: length of stay (LOS), disposition, three and 30-day return, and readmission rates. Only uncomplicated VOC visits due to SCD presenting to the EDOU were included in this study. Results: The average EDOU length of stay (LOS) for a patient with SCD during the emergency physician management period was 17 hours and 54 minutes; during the hospitalists physician management period, the average LOS was 18 hours and 23 minutes. Data were analyzed by patient disposition, and three and 30-day return. The admission rate during the emergency physician management period was statistically different than the admission rate during the hospitalists physican management period (16% versus 24.8% p<0.0001). There was no statistically significant difference in the three-day return rates and the 30-day return rates. Of the patients initially admitted to the hospital, the 30-day readmission rate under the emergency physician management period was 8.9% while it was 15.6% for the hospitalists physician management period (p<0.001) (not included in the figure). The EDOU LOS was not statistically different between groups (p¼.1853). Conclusion: While there are many factors that affect the rate of admission and ED utilization for uncomplicated VOC by patients with SCD. This research shows that there were both statistically significant and clinically significant differences in the hospital admission rates of patients with VOC during two time periods where care was managed by two different groups (emergency physicians and hospitalists physicians). Future research will examine the reasons for these differences.
Management Group Admittance Rates from an Emergency Department Observation Unit for Patients With Sickle Cell Disease
Mueller T, Sturgis L, Loeffler P, Kuchinski AM, Robert G, Lyon M/Augusta University, Augusta, GA
Study Objectives: Emergency department observation units (EDOUs) and sickle cell pathways (SCPs) allow for protocol-based rapid initiation of analgesic treatment for patients with sickle cell disease (SCD) during a vaso-occlusive crisis (VOC). There are many variations in the staffing and management of EDOUs. Our primary objective in this study was to compare hospital admission rates of patients treated in an EDOU for VOC due to SCD when managed by two different groups: Emergency physicians and hospitalists physician. In addition, we compared the three and 30-day return rates of patients when managed by these two different groups as a measure of premature discharge from the SCP.
Volume 68, no. 4s : October 2016
Annals of Emergency Medicine S33