Article

A comparison of perceived acceptable missed diagnosis rates for high-risk emergency medicine diagnoses: A brief report

additionally demonstrated in a recent narrative literature review article that ABM was not one of the modeling approaches for ED patient flow and crowding research [14]. However, each of the nine articles identi- fied here illustrates the value of applying ABM methodology to EM ad- dressing two main issues: ED operations and Infection control, although other subjects within EM may be appropriate for ABM as well. The obvious challenge for the implementation of ABM within EM is the lack of familiarity of EM researchers with this cutting-edge computa- tional approach. Didactic talks and workshops on ABM at local, regional, and national EM conferences will introduce a basic awareness of the prin- ciples of ABM to EM researchers and practitioners. In addition, cross-dis- ciplinary collaboration with ABM experts in other fields will benefit the

growth and development of ABM applications in EM research.

Another barrier is that as ABM has a symbiotic relationship with computing technology [15], practitioners of ABM require some degree of comfort with software skills [16,17]. In the future, simpler user inter- faces for ABM programs, should help to decrease the learning curve for the application of ABM to EM research.

In conclusions, our findings demonstrate that ABM can be of tremen- dous utility in studying ED operations, staffing patterns, patient triage algorithms, patient flow, wait times, and infection control strategies. This innovative approach to modeling can improve efficiency of ED op- erations and patient management as well as patient safety.

Conflicts of interest

All authors: no conflicts.

Funding sources

Dr. Hsieh is supported in part by an NIH award, K01AI100681, using a modeling approach to study HIV testing in Emergency Departments.

Jason M. Adleberg, BSE Christina L. Catlett, MD Richard E. Rothman, MD, PhD Yu-Hsiang Hsieh, MD, PhD*

Department of Emergency Medicine, Johns Hopkins University School of

Medicine, Baltimore, MD, United States

*Corresponding author at: Johns Hopkins University, Department of Emergency Medicine, 5801 Smith Avenue, Suite 3220 Davis Building,

Baltimore, MD 21209, United States.

E-mail address: [email protected] (Y-H. Hsieh).

Katie Lobner, MILS

William H Welch Medical Library, Johns Hopkins University School of

Medicine, Baltimore, MD, United States

27 September 2016

https://doi.org/10.1016/j.ajem.2017.06.021

References

  1. Wooldridge M. Agent-based software engineering. IEE Proc Software Engineering 1997;144(1):26-37.
  2. Epstein J, Axtell R. Growing artificial societies: social science from the bottom up. Washington, DC: Brookings Institution Press & MIT Press; 1996.
  3. Epstein J. Generative social science: studies in agent-based computational modeling. Princeton, New Jersey: Princeton University Press; 2007.
  4. Jones S, Evans R. An agent based simulation tool for scheduling emergency depart- ment physicians. AMIA Annu Symp Proc 2008;2008:338-42.
  5. Rahmat M, Annamalai M, Halim S, Ahmad R. Agent-based modelling and simulation

    of emergency department re-triage. 2013 IEEE business engineering and industrial applications colloquium (BEIAC); 2013. p. 219-24 Langkawi, Malaysia.

    Taboada M, Cabrera E, Iglesias M, Epelde F, Luque E. An agent-based decision support sys-

    tem for hospital emergency departments. Procedia Computer Science 2011;4:1870-9.

    Wang L. An agent-based simulation for workflow in emergency department. 2009 IEEE sys- tems and information engineering design symposium; 2009. p. 19-23 Charlottesville, VA.

  6. Laskowski M, Demianyk B, Friesen M, McLeod R. Uncertainties inherent in RFID Tracking systems in an emergency department. 2010 IEEE workshop on health care management (WHCM). Venice: IEEE; 2010. p. 1-6.
  7. Laskowski M, McLeod R, Friesen M, Podaima B, Alfa A. Models of emergency depart- ments for reducing patient Waiting times. PLoS One 2009;4(7):e6127.
  8. Neighbour R, Oppenheimer L, Mukhi S, Friesen M, McLeod R. Agent based modeling of “crowd informing” as a means of load balancing at emergency departments. On- line J Public Health Inform 2010;2(3):e4.
  9. Laskowski M, Demianyk B, Witt J, Mukhi S, Friesen M, McLeod R. Agent-based modeling of the spread of influenza-like illness in an emergency department: a sim- ulation study. IEEE Trans Inf Technol Biomed 2011;15(6):877-89.
  10. Aleman D, Wibisono T, Schwartz B. Accounting for individual behaviors in a pan- demic disease spread model. Winter simulation conference. Austin, TX: winter sim- ulation conference; 2009. p. 1977-85.
  11. Wilensky U. NetLogo. Evanston, IL: Center for Connected Learning and Computer-

    Based Modeling, Northwestern University; 1999.

    Wiler J, Griffey R, Olsen T. Review of modeling approaches for emergency depart- ment patient flow and crowding research. Acad Emerg Med 2011;18(12):1371-9.

  12. Gilbert N, Bankes S. Platforms and methods for agent-based modeling. Proc Natl

    Acad Sci U S A 2002;99(Suppl. 3):7197-8.

    Minar N, Burkhart R, Langton C, Askenazi M. The swarm simulation system: a toolkit for building multi-agent simulations. Santa Fe, NM: Santa Fe Institute; 1996.

  13. Ropella G, Railsback S, Jackson S. Software engineering considerations for individual- based models. Nat Resour Model 2002;15(1):5-22.

    A comparison of perceived acceptable missed

    diagnosis rates for high-risk emergency medicine diagnoses: A brief report?,??

    Current practice in emergency medicine (EM) involves ruling out low-probability, high-risk diagnoses in patients presenting with com- mon chief complaints. This approach frequently differs from that taken by non-EM providers and can result in contrasting views on an appropriate diagnostic evaluation and disposition. Medical liability re- mains a concern with many high-risk emergency medicine diagnoses [1]. Therefore, it is not surprising that the risk tolerance for a missed di- agnosis is known to vary widely among emergency physicians, and re- sults in a wide range of practice patterns [2,3]. The purpose of this study is to compare the perceived acceptable missed diagnosis rate for several life-threatening or common emergency medicine diagnoses be- tween emergency medicine providers, non-emergency medicine pro- viders, and patients and their families.

    We conducted a prospective cross-sectional survey study at an aca- demic, tertiary-care medical center with an annual census of N 82,000. Subjects were divided into three groups: 1) EM Providers (EMP), from EM-specialized advanced practice practitioners (APPs) to EM residents and attendings; 2) non-EM providers (non-EMP), which included APPs, residents and attendings, at varying levels of training, practicing in Family Medicine or internal medicine and its subspecialties; and 3) patients, who presented to the ED for evaluation, their family members, or their accompanying acquaintances (PFA). A convenience sample of participants were randomly approached in the ED waiting room, at res- ident conferences, faculty meetings, and administrative areas. Exclusion criteria included an age b 18 years old, prisoners, military basic trainees, or the inability to read or write in English.

    The survey contained demographic information and eight clinical sce- narios: myocardial infarction (MI), pulmonary embolism (PE), Ruptured aortic aneurysm (AAA), cerebrovascular accident (CVA), appendicitis, sub- arachnoid hemorrhage (SAH), meningitis, and Ectopic pregnancy (see Ap- pendix 1). Participants were asked to determine, “how often is it acceptable for a healthcare provider to miss the possibly serious or life- threatening diagnosis” by selecting one of five fixed percentages (10%, 5%, 1%, 0.1% or <=0.0001%). The survey was developed and piloted to account

    ? Abstract presentation at the following scientific meetings: American College of Emergency Physicians Scientific Assembly, Chicago, IL, Oct 2014.

    ?? Disclaimer: The view(s) expressed herein are those of the author(s) and do not

    reflect the official policy or position of Brooke Army Medical Center, the U.S. Army Medical Department, the U.S. Army Office of the Surgeon General, the Department of the Army and Department of Defense or the U.S. Government.

    Table 1

    Most frequently chosen acceptable missed diagnosis rate (mode) by survey question and group and comparison of responses between groups by survey question (p-values).

    Question – Diagnosis Mode Comparison between groups

    (p-value b 0.05 in bold)

    EMP (% of total)

    Non-EMP (% of total)

    PFA (% of total)

    EMP vs. non-EMP

    EMP vs. PFA

    Non-EMP vs. PFA

    #1 – MI

    1% (44)

    0.1% (38)

    <= 0.0001% (53)

    0.0005

    0.0025

    0.2341

    #2 – PE

    1% (49)

    0.1% (38)

    <= 0.0001% (43)

    b0.0001

    0.0001

    0.1691

    #3 – AAA

    1% (40)

    0.1% (44)

    <= 0.0001% (37)

    0.028

    0.0037

    0.8902

    #4 – CVA

    1% (41)

    0.1% (41)

    <= 0.0001% (47)

    b0.0001

    0.0031

    0.1118

    #5 – Appendicitis

    1% (38)

    0.1% (44)

    <= 0.0001% (50)

    b0.0001

    0.0005

    0.0109

    #6 – SAH

    1% (44)

    1% (34)

    <= 0.0001% (34)

    0.0139

    0.0171

    0.8399

    #7 – Meningitis

    1% (48)

    0.1% (40)

    <= 0.0001% (35)

    0.0074

    0.0209

    0.4921

    #8 – Ectopic pregnancy

    0.1% (44)

    b0.0001% (40)

    <= 0.0001% (48)

    0.1446

    0.0918

    0.9731

    for a broad range of risk tolerance, while balancing it with percentages that approximated real-world missed diagnosis rates for EM diagnoses.

    A total of 231 out of 240 distributed surveys were completed for a study response rate of 96%. In total, there were 80 EMP (45% attending physicians), 69 non-EMP (24.2% attending physicians), and 77 PFA (49.4% patients) completed surveys included for data analysis. The aver- age age was 37.5 years (EMP 33.7 years old, non-EMP 37.3 years old, PFA 41.9 years old) consisting of 55.3% male participants. When com- paring groups using the Wilcoxon method, the EMP chose a significantly higher acceptable missed diagnosis rate when compared to non-EMP or PFA responses, for all clinical scenarios except ectopic pregnancy. The EMP most frequently chose 1% as an acceptable missed diagnosis rate (except for ectopics), non-EMP most frequently chose 0.1% (except for SAH and ectopics), and PFAs most frequently chose <= 0.0001%. Non- EMPs and PFAs were not significantly different from each other (except for appendicitis), who generally chose lower acceptable missed diagno- sis rates than the EMP. (See Table 1 and Graphs 1 and 2.)

    The main finding of this survey study is EMPs tend to choose a higher acceptable missed diagnosis rate for low-probability, high-risk diagno- ses, when compared to non-EMPs and PFAs. In contrast, there is no dis- cernible difference in the acceptable missed diagnosis rate between non-EMPs and PFAs, except when it comes to a relatively less serious di- agnosis, such as appendicitis. The most frequently chosen acceptable missed diagnosis rates between groups is in clear juxtaposition to the anecdotal experiences and actions observed in EDs. In our institution, EMPs are often perceived as too conservative in their management plans (e.g., admit patients unnecessarily) relative to non-EMPs (e.g., dis- charge home Low-risk chest pain patients that we consult for admis- sion), however, our data demonstrates a contradiction in our perception of non-EMP risk tolerance.

    An oft-quoted study on EMPs reports AMI is mistakenly discharged home from the ED 2.1% of the time, which may explain why they most frequently chose 1% as an acceptable missed diagnosis rate [4]. In contrast, non-EMPs and PFAs chose 0.1% and b 0.0001%, respectively. These tendencies suggest that a greater proportion of chest pain pa- tients would need to be admitted to achieve these exponential reduc- tions in acceptable missed diagnosis rates, perceived by non-EMPs and PFAs, over what is occurring with current EMP practice patterns. These differences between the EMPs and the other groups surveyed have many potential implications for clinical practice, as the likelihood of a given disease process often affects evaluation priorities.

    In summary, EMPs are more likely to consider higher missed diagno- sis rates acceptable than non-emergency medicine providers and pa- tients, their family or their acquaintances. These data provide some insight regarding the different frame of reference concerning risk toler- ance by EMPs, non-EMPs, and PFAs, for several life-threatening or com- mon emergency medicine diagnoses.

    Funding

    None.

    Appendix A. Survey Questions

    Introduction

    You are being asked to voluntarily participate in an anonymous sur- vey study to evaluate what you consider to be an “acceptable miss rate” for a number of potentially life-threatening or permanently debilitating diagnoses or medical problems seen in emergency departments across the United States. This survey should only take 5-10 min.

    For example, if you or a close family member presented to an

    emergency department with common Cold symptoms (i.e., runny nose, cough, sore throat, fever), but your health care provider missed a potentially life-threatening diagnosis of a blood infection (sepsis), whether or not a bad outcome occurs (i.e., Prolonged hospitalization, permanent disability or death), how often would you consider a missed diagnosis to be acceptable and just due to “bad luck”?

    Note: There are no correct answers to any of the questions below.

    If you or a close family member presented to an emergency department with chest pain, how often is it acceptable for a healthcare provider to miss a possibly serious heart attack (myocardial infarction)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  14. 1 in 20 patients (5%)
  15. 1 in 100 patients (1%)
  16. 1 in 1000 patients (0.1%)
  17. 1 in 1,000,000 patients or less (<= 0.0001%)
  18. If you or a close family member presented to an emergency depart- ment with shortness of breath, how often is it acceptable for a healthcare provider to miss a possibly serious blood clot in the lungs (pulmonary embolism)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  19. 1 in 20 patients (5%)
  20. 1 in 100 patients (1%)
  21. 1 in 1000 patients (0.1%)
  22. 1 in 1000,000 patients or less (<=0.0001%)
  23. If you or a close family member presented to an emergency depart- ment with low back pain, how often is it acceptable for a healthcare provider to miss a possibly serious leaking large blood vessel (rup- tured aortic aneurysm)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  24. 1 in 20 patients (5%)
  25. 1 in 100 patients (1%)

    Graph 1. survey responses: Questions #1-4.

    Graph 2. Survey responses: Questions #5-8.

    1 in 1000 patients (0.1%)

  26. 1 in 1000,000 patients or less (<= 0.0001%)

    If you or a close family member presented to an emergency department with numbness in the arm, how often is it acceptable for a healthcare provider to miss a possibly serious stroke (cerebrovascular accident)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  27. 1 in 20 patients (5%)
  28. 1 in 100 patients (1%)
  29. 1 in 1000 patients (0.1%)
  30. 1 in 1000,000 patients or less (<= 0.0001%)
  31. If you or a close family member presented to an emergency depart- ment with abdominal (belly) pain, how often is it acceptable for a healthcare provider to miss a possibly serious intestinal infection of the appendix (appendicitis)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  32. 1 in 20 patients (5%)
  33. 1 in 100 patients (1%)
  34. 1 in 1000 patients (0.1%)
  35. 1 in 1000,000 patients or less (<= 0.0001%)
  36. If you or a close family member presented to an emergency depart- ment with a headache, how often is it acceptable for a healthcare pro- vider to miss a possibly serious bleeding blood vessel in the brain (berry aneurysm)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  37. 1 in 20 patients (5%)
  38. 1 in 100 patients (1%)
  39. 1 in 1000 patients (0.1%)
  40. 1 in 1000,000 patients or less (<= 0.0001%)
  41. If you or a close family member presented to an emergency department with a flu-like illness, how often is it acceptable for a healthcare provider to miss a possibly serious brain covering infection (meningitis)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  42. 1 in 20 patients (5%)
  43. 1 in 100 patients (1%)
  44. 1 in 1000 patients (0.1%)
  45. 1 in 1000,000 patients or less (<= 0.0001%)
  46. If you or a close family member presented to an emergency depart- ment with abdominal (belly) pain and Vaginal bleeding, how often is it acceptable for a healthcare provider to miss a possibly serious bleeding tubal pregnancy (ruptured ectopic pregnancy)? Due to the missed diagnosis, you might become disabled, need to stay in the hospital for a long time, or even die (circle only one answer).

    1 in 10 patients (10%)

  47. 1 in 20 patients (5%)
  48. 1 in 100 patients (1%)
  49. 1 in 1000 patients (0.1%)
  50. 1 in 1000,000 patients or less (<= 0.0001%)
  51. Have you or a close family member ever experienced a misdiagnosis of a serious medical illness?

    No

  52. Yes

    Eric J. Chin, MD

    Department of Emergency Medicine, San Antonio Military Medical Center,

    JBSA-Ft. Sam Houston, TX, USA Corresponding author at: San Antonio Military Medical Center, Dept. of Emergency Medicine, 3551 Roger Brooke Drive, JBSA-Ft Sam Houston,

    TX 78234, USA.

    E-mail address: [email protected].

    Andrew Bloom, MD

    Department of Emergency Services, Irwin Army Community Hospital, Ft.

    Riley, KS, USA

    Andrew Thompson, MD

    Spartanburg Medical Center, Spartanburg, SC, USA

    6 January 2017

    https://doi.org/10.1016/j.ajem.2017.06.036

    References

    Brown TW, McCarthy ML, Kelen GD, et al. An epidemiologic study of closed emergen- cy department malpractice claims in a national database of physician malpractice in- surers. Acad Emerg Med May 2010;17(5):553-60.

  53. Brown TB, Cofield SS, Iyer A, et al. Assessment of risk tolerance for adverse events in emer- gency department chest pain patients: a pilot study. J Emerg Med 2010;39(2):247-52.
  54. Katz DA, Williams GC, Brown RL, et al. Emergency physicians’ fear of malpractice in evaluating patients with possible acute cardiac ischemia. Ann Emerg Med Dec 2005;46(6):525-33.
  55. Pope JH, Aufderheide TP, Ruthazer R, et al. missed diagnoses of acute cardiac ischemia in the emergency department. NEJM 2000;342:1163-70.

    Capsaicin topical in emergency department treatment of Cannabinoid hyperemesis syndrome

    Marijuana is reported as the most commonly used illicit drug in the United States [1]. With the increasing rates of marijuana use, the recognition of a new clinical condition has followed. Cannabinoid hyperemesis syndrome (CHS) is a clinical condition characterized by chronic cannabis use, cyclical nausea and vomiting, and compulsive bathing in hot water [2]. The condition is separated into three phases: prodrome, vomiting, and recovery. The prodromal phase of CHS is characterized by acute nausea and abdominal discomfort. The vomiting phase is character- ized by severe nausea, vomiting, and retching that can occur up to five times per hour [2]. During this phase patients may exhibit compulsive bath- ing in hot water, which patients sometimes find to relieve their nausea and vomiting. Patients may visit the hospital after they have run out of hot water and can no longer control their symptoms [3]. While in the hospital, imaging and blood work are typically conducted and most of the time are unrevealing. Typical anti-emetics are given to these patients, usually with little to no relief. The recovery phase begins when the patients stops con- suming cannabis. Within one week of cessation, symptoms of nausea, vomiting, and abdominal pain will significantly decrease [2].

    Although CHS is associated with cannabis use, it is unknown if cannabi- noids are the cause. The mechanism of CHS remains unknown. One theory is that it is due to an accumulation of tetrahydrocannabinol (THC) in heavy marijuana users. THC, the active component of most types of marijuana, binds to cannabinoid type 1 (CB1) receptors in the GI tract resulting in gastroparesis and severe nausea and vomiting. With chronic cannabis con- sumption the CB1 receptors can become sensitized, leading to the pro-emet- ic CB1 activity in the GI tract which overrides the anti-emetic properties of CB1 receptors in the brain [4]. Other mechanisms proposed include down- regulation of CB1 receptors. In the presence of reduced functionality of CB1 receptors the agonistic nature of THC can turn into antagonistic, causing an emetic effect rather than the known anti-emetic effect of marijuana [5].

Leave a Reply

Your email address will not be published. Required fields are marked *