Statistics in the Emergency Department

I haven't been active in this site for months since I started working on some work-related stuffs; add the post-graduate courses and extra curricular activities to that and *poof!* no time for blogging. Right now, I am still busy catching up with backlogs from my 2 post-grad courses, Statistical Methods Applied in Nursing and Advanced Adult Health Nursing. Let me share one of my forum posts about why Stat is important in my field of practice.

The Emergency Department (ED) has always been the most unpredictable area in any hospital worldwide. There are times that the patient influx can be overwhelming especially during the peak hours; and there are times that we can just sit back and relax trying to forget that an hour ago, we just had a catastrophic battle with blood and gore in the ED. Although we say that the emergency room can be horribly busy, that does not apply all the time. We can be working on very simple cases ranging from minor cuts and fever management to the management of severe burns and cardiac arrest. No matter how we can be ‘benign’ or ‘toxic’ with patient management, we are always guided by clinical protocols, guidelines and policies. And these policies were developed (and revised) from continuous analysis of trends and patterns obtained from different forms and sources of data within the institution. Let me give examples of how statistics could be of great help in improving the operations in the ED. 

Data Analysis. The ED may simply want to obtain and analyze data that would reflect its overall performance and identify trends. These can be one (but not limited to) of the following:

  • Handover times of ED nurses to ICU, to the general ward, to the Cardiac Catheterization Laboratory, etc.,
  • Frequency of needle prick injuries sustained by nurses,
  • Blood pressure measurements of ED patients above 40 years of age,
  • Number of ED returns within 24 hours,
  • Number of RTAs/MVAs according to month, year, victim (gender, age), etc.,
  • Percent of ED visits that require transfer to a different hospital,
  • Percent of ED visits with patient seen in fewer than 15 minutes,
  • Percent of ED visits that require hospital admission,
  • Success rate of IV cannulation by ED nurses with 2 years, 4 years, and 6 years of experience,
  • Frequency of ED visits of “this suspected Demerol-addicted patient”
  • Number of Jane Doe/John Doe sent to the ED,
  • Correlation between income levels and tendency to seek treatment in EDs,
  • Five, Ten, Fifteen years analysis of ED census, and a lot more.

The 2009 AH1N1 scare all over the world could have been a very good subject for a nurse clinician to work on for research. Different aspects of AH1N1 care such as patients’ responses and initial reactions upon the onset of symptoms could have been analyzed, the symptomatology could have been clustered to different types of data. At the peak of the AH1N1 scare, I had some consecutive duty days wherein I was exclusively assigned to do bedside management to suspected AH1N1 carriers. There were a number of us who became “AH1N1 nurses” that time and fortunately none of us got the dreaded virus (or maybe nobody reported or was tested for it; well, I had flu-like symptoms that time and was given 4 days sick leave! haha. ). I was thinking, a retroactive study of the incidence of disease transmission to isolation nurses could have been a decent research study! This, the things I mentioned earlier, and more could be potential research topics that need statistics to come up with trends and compare against some hypotheses.

Triage Acuity Assignment and Waiting Times. As patients come into the ED, the ones who need care most urgently receive it first. This is one of the areas that the Triage Nurse has to address to regularly – backlogs. Waiting times are always a concern of the ED especially when the patients’ remark, “long waiting time,” is reflected in EVERY patient satisfaction survey. In some cases, however the ED attempts to educate the clients about triage acuities (providing information materials about triage), these patients may still complain of the same problem. The management should then take a good look at other variables that may affect the ED clients’ satisfaction rate.

Hospital’s Adherence to Established Protocols. In an institution’s attempt to do clinical benchmarking, the administration may want to study the ED’s adherence to the American Heart Association’s suggested door to ECG time (10 minutes) for clients with chest pain. The same study can also be done for door to CT Scan and door to needle (thrombolytic) time in suspected stroke patients.

Bed Management Cycle. Patients triaged with Acuity Level 4 or 5 may not need to be placed on a bed for management. Often times, these clients may just need a simple consultation for simple prescription medications and this happens frequently when Out Patient Department offices are closed. And for some reasons, about 60% (my rough estimate based on experience – needs validation, haha.
) of ED visits fall to this triage acuity level, hence, the need for some institutions to put up fast-track rooms in the ED. Analysis of bed and equipment utilization for Urgent, Emergent, and Resuscitation rooms should always be considered especially in EDs with minimal bed capacities but with high patient census.

Staffing at the ED. Although EDs operate 24 hours a day, staffing levels may be varied in attempts to mirror patient volumes. For instance, in my previous institution in the Philippines (30-bed capacity Emergency Department) it has been established that the start of the rise of the census curve before reaching the peak hours of patient visits is at 1000H and the most number of patients are handled by the evening shift (1400H-2200H). The curve gradually decreases as the night time comes until the retirement hours begin. Usually, 8 bedside nurses are on duty at the start of the morning shift, and then two more nurses should arrive by 1000H to do a 12-hour shift. The addition of the two staff nurses should cover the patient volume during the peak hours. The information obtained from simple analysis of trends led to the establishment of a minimal staffing pattern for the institution’s adequate delivery of service. With the use of simple statistics, the staffing in the ED can indeed be managed.

Creation of Protocols and Clinical Pathways. Stomach pain, headache, chest pain, and fever are among the most common reasons why people visit the ED. However, this may not be true to other institutions in a specific geographical boundary. Road traffic accidents may top the list for other EDs while fever could be the topnotcher for other hospitals. Simple statistical analysis could be used to classify chief complaints and create a clinical pathway to standardize the most basic management of the common cases.

Seasonal Trending. At the start of the rainy season in the Philippines, July or August, suspected leptospirosis cases are expected to increase; patients coming in for flu-like symptoms may also rise. In December, during the holiday season, food related cases such as indigestion and food poisoning are expected to go up. Through statistics, the pattern and trends of patient influx during a specified month/year could be analyzed to foresee the overall need for timely resource management.

Identification of trends and patterns of the emergency department’s operation is necessary so that better health care planning and service provision can be implemented.


Kristin Peña said…
Hello. I am a medical student and I am looking for a topic for our Master's Degree on Public Administration, Major in Health Emergency and Disaster Management. I really found your article helpful. It gave me some insight about my topic of interest which was about Emergency Department Crowding and your statement, "how statistics could be of great help in improving the operations in the ED," gave me that idea. :) Thanks.

Popular posts from this blog

Prometric Exam / Saudi Council Exam

DepEd Authentication of High School Documents

A Filipino Migrant's View of Living in New Zealand