Add-on cases in the endoscopy unit: Factors that affect volume

Department of Medicine, University of British Columbia, St Paul’s Hospital,Vancouver, British Columbia Correspondence: Dr Robert Enns, St Paul’s Hospital, University of British Columbia, 770-1190 Hornby Street, Vancouver, British Columbia V6Z 2K5. Telephone 604-688-6332, fax 604-689-2004, e-mail renns@interchange.ubc.ca Received for publication August 12, 2008. Accepted September 25, 2008 O the past several decades, endoscopic procedures have quickly evolved, expanding in both diagnostic and therapeutic avenues (1). In 2002, 1444 Canadian physicians performed a minimum of 100 procedures each, accounting for a sum of nearly one million procedures (2). The total number of colonoscopies reported was 523,224, while gastroscopies totaled slightly fewer at 392,568. Physicians who performed less than 100 procedures per year were not included in the report; therefore, these totals are likely an underestimate of the true procedure totals. originAl Article

At St Paul's Hospital, Vancouver, British Columbia, a team of eight gastroenterologists complete approximately 10,000 gastrointestinal endoscopic procedures annually. Most of these procedures are electively scheduled, with patients (as outpatients) being discharged home the same day, typically within 1 h to 2 h of their procedure. Add-on procedures are usually emergency procedures and performed on inpatients. They require a significant allocation of endoscopy resources because they are urgent, patients are not readily discharged, tend to use more equipment and often involve endoscopic hemostasis. In addition, patients tend to be more acutely ill with significant comorbid disease and, therefore, higher risk. The addition of these patients to a typical endoscopy slate increase the workload and require careful planning to ensure that the consumption of resources remain within budgetary limits (3).
Because staffing is readily available during normal work hours, regularly scheduled cases performed during peak operational times in the gastroenterology (GI) clinic typically do not involve overtime requests which involve extraneous payroll costs. To specifically target overtime and resource costs, and properly allocate resources, it is of critical importance to actually recognize situations in which add-on cases are most prevalent. Efficiency in the endoscopy unit has been studied in many different ways (4-7). However, very few studies have addressed the issue of add-on cases (8)(9)(10). In an effort to determine factors that correlate with a high volume of add-on emergency cases, we retrospectively reviewed eight months of add-on and elective case data from a tertiary care, hospitalbased endoscopy unit.

METHODS
A retrospective office chart review was performed on all nonemergent gastrointestinal endoscopy cases performed between September 2006 and May 2007 in the GI clinic at St. Paul's Hospital. For the purpose of the present study, the focus was on data from add-on and electively booked cases. Specifically, all gastroscopies, colonoscopies, and flexible sigmoidoscopies were reviewed. The day of the week and month a procedure was performed, type of procedure and physician were recorded. Emergency add-on cases were then compared with a randomly selected group of electively booked cases (during the same time period) to determine if differences in demographics or procedure type could be determined. It should be noted that emergency add-on procedures completed during the weekend were not assessed in the present study because our focus was the interaction of add-on cases with elective cases during the week.
The null hypothesis to compare physician add-on cases was that all physicians were adding on an equal number of cases adjusted by call day. The Kruskal-Wallis rank sum test was used to compare the equality of proportions.
Add-on procedures were more likely to be gastroscopies than in the elective cases (OR 2.7; 95% CI 1.8 to 4.3; P<0.0001) and less likely to be colonoscopies (OR 0.24; 95%CI 0.15 to 0.38; P<0.0001).
The day of the week varied in the frequency of add-on cases. Monday proved to be the busiest day for cases being added to the normal slate (OR 1.7; 95% CI 1.0 to 2.28; P=0.03). Conversely, Fridays revealed significantly fewer add-on procedures (OR 0.31; OR 0.16 to 0.57; P=0.0001).
Variation was also noted with respect to time of the year, and more specifically, by month. Statistically, September was the least busiest month (40 cases) of the year for add-on procedures (OR 0.31 95% CI 0.11 to 0.78; P=0.0006) in the time frame studied, while the month of March had a total of 109 additional patients added.

add-on cases in the endoscopy unit
Finally, the number of cases added on among eight physicians over their call period was compared. Statistically, the number of add-on cases for all eight physicians was not equal (c 2 using H statistic for ties = 18.324 with seven degrees of freedom; P=0.0106).

DISCUSSION
Within the current working framework of the GI clinic at St Paul's Hospital, only absolute emergent cases are completed on weekends, leaving additional cases (which are still emergent) that arise over the weekend to be completed the following week. This practice may not be ideal and may increase the length of hospital stay (thereby increasing costs allocated to that individual patient) (8,12). However, this is common practice in many Canadian centres. Nonetheless, cases from Saturday and Sunday often will be stabilized with intravenous fluid resuscitation and endoscopically evaluated on Monday. Although the number of add-on cases were similar for Monday, Tuesday and Wednesday, when long weekends are accounted for (cases are added on Tuesday instead of Monday and add these cases to the usual first working day of the week), Monday tended to have more add-on cases than any other day. Interestingly, the early part of the week was heavy with add-on cases; consequently, Fridays were noted to have the fewest number of add-on procedures. Because upper gastrointestinal bleeding is the most common urgent condition, esophagogastroduodenoscopy was performed most often for these add-on procedures. Conversely, 60% of the elective cases (n=442) were colonoscopies. This highlights the fact that these procedures were used primarily for investigation of colonic abnormalities and/or pre-emptive screening procedures for cancer, which has become common for outpatients.
The reason for Fridays having fewer add-on cases is open to speculation. Possibly, a change of the physician on call that day may result in most cases being completed earlier in the week. The physician finishing his call schedule on Friday at 08:00 likely makes special effort on the days before to complete his/her cases and thereby adds more cases to the previous days to avoid excessive 'handover' of cases. Furthermore, investigation of this factor revealed one 'outlier' with respect to the frequency of add-on cases, which may be due to a number of reasons. However, this only highlights the many possible factors that influence the number of add-on cases in a GI clinic.
Because St Paul's Hospital is a teaching hospital, residents and GI fellows are involved in virtually all consultations. At times, this may affect the numbers of cases seen each week because some advanced trainees may be more adept at 'steering' cases away (ie, outpatient appointments). Alternatively, they may aggressively advocate for endoscopic assessment and, thus, increase the number of endoscopic procedures. This is clearly a confounder that could account for some week to week variation that was not accounted for in the present study. Because a GI fellow is present all year, assessing their tendencies toward or against aggressive endoscopy practices is difficult.
Seasonal changes relative to which months are the busiest for add-on cases revealed September as the least busiest month. Multiple hypotheses may begin to explain this anomaly. However, patients returning to work or school may prevent them from seeking medical attention for a given gastrointestinal ailment until after their lives settle down (eg, October). It is possible that this hypothesis fits because there was an increase in add-on procedures slated in October.
A limitation of the present study is the fact that the entire year was not assessed and therefore there were several months that were not evaluated. This was a logistical issue because access to data regarding the several absent months was not possible on the database used. Therefore, we cannot comment on each and every month and simply state that the number of emergent endoscopic assessments in each month were not equivalent in the time period studied.
There also were statistically significant differences in the number of cases added on by specific staff; despite the fact that the environment is presumably relatively stable. Although this may be partially accounted for by fluctuations in the number of in-hospital consultations and variability in gastrointestinal bleeding, the fact that at least six weeks of call was reviewed for each staff member over a time period of one year would usually have normalized this variation. There are likely different