07/2011 – 06/2014
- ED Dashboard — Developed a set of daily dashboards sends to ED directors and hospital executives detailing Emergency Department metrics including performance measures, patient flow, and operations
- Reporting System — Developed a robust and versatile reporting system that sends daily reports to various departments and personnel aiding in patient safety, follow-up, and recruitment
- Physician Communication — Developed software to contact physicians through email upon patient arrival to and disposition from the ED by supplementing our communication with PMDs and in-house physicians
- Procedure Logger — Created software to automatically sync daily resident procedures with New Innovations to allow for accurate logging and reporting
- Meaningful Use and eMeasures — Aided the hospital’s organizational performance section with extracting, compiling, and reporting data from our EMR for clinical quality measures
05/2007 – 04/2011
2005 – 2006
08/1998 – 12/2001
7/2014 – Present
Arlington Emergency Medical Associates (AEMA) is an independent, democratic group with 22 clinically practicing partners providing exceptional level of care and attention and have done so continually for the past three decades. AEMA physicians and the exceptional ED staff work together to provide high-quality, comprehensive care.
06/2006 – 03/2007
- Delivered excellent first and second tier support to international and domestic clients. Focused on providing thorough and
direct resolutions to client issues on a daily basis.
- Provided technical skills and expertise to Client Support department in the form of growing knowledge of Verizon binary implementation, Perl scripts written to minimize engineering time spent on reports and gathering information from logs.
- Designed and implemented company wide projects impacting the Technical Support department. Collaborated and engaged with other technical and non-technical departments to resolve customer issues.
- Lead Service representative for all phases of development of SMS projects involving AT&T, Verizon, and T-Mobile.
05/2002 – 05/2006
- Designed network architecture and security policies for several consulting groups and marketing firms handling several million website visitors a day. Focused on providing excellent border security through firewall and synflood protection devices, operating system hardening, and access level protection.
- Created incident response policies and guidelines, developed security tools and implemented immediate network support escalation in face of attacks. Worked closely with clients to ensure a secure computing environment including remote access as well as host and application security. Maintained the physical security of networks and hosts through managed services.
- Attained broad knowledge of networking equipment such as the Netscreen 5-200 firewalls, Ultramonkey load balanced
solution, Radware WSD-Pro load balancers, Cisco Catalyst 2900/4500 switches, Alpine 3800 series switches, Compaq Proliant 1850/3000, Netscreen IDP 1000 (certificated), and Netscreen Secure Access 5000 (certificated).
- Experienced system administrator of Debian and Red Hat Linux, FreeBSD, and Solaris 8/9 and related system services including but not limited to Apache Web Server, PostgreSQL database, Qmail and Postfix mail servers, Bind Name Server,
and Netfilter packet filtering framework.
Board Certified, Expiration 12/31/2016
2011 – Present
Registered with ClinicalTrials.gov — NCT01835262
Published in the American Journal of Emergency Medicine (2014)
Ketamine infusion is a potentially enticing alternative to ketamine push for having less incidence of adverse effects of ketamine while maintaining ketamine’s proven analgesic efficacy. So far, ketamine has shown great potential, especially for patients who do not respond to opioid medications. At this time, however, more research is needed before implementing widespread use of subdissociative dose ketamine infusion for acute pain in the ED. Further study is also necessary for ketamine for pain control in hemodynamically unstable patients for whom opioids pose a danger of hemodynamic decompensation.
PMID: 25447602 [PubMed – as supplied by publisher]
Moderated Poster Presentation — Council of Emergency Medicine Residency Directors Academic Assembly in New Orleans, LA and Society for Academic Emergency Medicine 2014 in Dallas, TX
Emergency medicine residents document procedures they perform to paper or web-based logs, but no studies have utilized the electronic medical record (EMR) to automatically record procedures in a web-based log to assess the number of captured procedures.
To assess the impact of an automated procedure logging system on the number of procedures documented in an emergency medicine residency program.
We conducted a single center retrospective study at an urban community teaching hospital with an annual census of 120,000 assessing the quantity of 17 ACGME specified key index procedures performed during the 2011 and 2012 calendar years. A Perl application was written to access the EMR on a daily basis, retrieve procedures performed by residents, populate a comma-delimited file with procedures, and using secure shell protocol to securely transfer the file to the data warehouse. The program application was introduced January 1, 2012. The study collected the total number of procedures documented, procedures per residency class, and percent change per procedure.
There was no statistically significant difference in the number of patients seen per resident or departmental acuity in the pre- and post-intervention period. Those procedures requiring any additional effort to document (patient resuscitations and ultrasonography) showed an 11% decrease while completely automated logs showed a 32% increase in procedure numbers with a net difference of -994 and +1178, respectively. The increase in procedure numbers auto-logged was evenly distributed among the different residency classes: R1 29%, R2 29% and R3 42%.
Automating the documentation significantly increased the number of procedures logged, likely reflecting a more accurate capture of the number of procedures being performed.
04/2014 – Present
Registered with ClinicalTrials.gov — NCT02078492
2/2014 (in manuscript)
Poster Presentation — American Academy of Emergency Medicine 2014; New York City, NY and Florida AAEM Scientific Assembly in Miami Beach, FL
05/2013 – Present (in manuscript)
Oral Presentation — Society for Academic Emergency Medicine 2013; Atlanta, GA
Poster presentation — Society for Academic Emergency Medicine 2011 in Boston, MA
The EDWIN and NEDOCS models were created to quantify crowding within the emergency department. In addition to the calculated scores, the models provide clear insight into the departmental staffing, bed occupancy, and other components that factor into ED crowding.
To assess emergency department crowding utilizing the EDWIN and NEDOCS models. The hypothesis of this study was that the models would not accurately measure crowding due to observed and statistical assumptions in the respective formulas.
Calculate hourly EDWIN and NEDOCS scores retrospectively from January 1, 2009 to June 30, 2010 and collect data related to each of the formulas used by the respective models.
There were 13,104 hourly scores calculated for both the EDWIN and NEDOCS models. The EDWIN model showed that the ED was active 8833 hours (67.4%), busy 1860 hours (14.2%), and crowded 2411 hours (18.4%) during the entire study period. The NEDOCS model showed that the ED was very busy 3976 hours (30.3%), overcrowded 3771 hours (28.8%), dangerous 2102 hours (16.0%), and disaster 1534 hours (11.7%). Overall, the ED was overcrowded, in dangerous, and disaster levels 7407 hours (56.5%) of the time. The EDWIN and NEDOCS scores correlated weakly during the entire study period with the correlation coefficient calculated as 0.29.
While both the EDWIN and NEDOCS represent valiant attempts to objectively quantify emergency department crowding, there are limitations of both models that make general applicability unrealistic at this time. EDWIN breaks down when the number of admitted patients equals or exceeds the number of available ED beds [Na * (Bt – Ba)], producing a denominator of 0 or a negative number in one of the terms of the formula. In contrast, the NEDOCS model becomes dominated by the term, [85.8 * (Pt / Be)], when the number of total patients (Pt) exceeds the number of ED beds (Be). This results in a relatively high contribution toward the overall calculated score. It is our belief that future models will need to overcome the assumptions within each of the model and instead incorporate advanced concepts in queueing theory that can better analyze the overall input-throughput-output of patient flow.
08/29/2011 – 08/31/2011
- Assyrian Foundation of America Student Scholarship
- UC Berkeley Lisa Rist Memorial Scholarship
- UC Berkely Rodkey Scholarship
- Kaiser Permanente Volunteer of the Year and Centennial Scholarship
- Sam Walton Community Leader Scholarship
- Bank of America Achievement Award
- Go-For-It Scholarship (Santa Clara County)
Associate member of the American Academy of Emergency Medicine