AIM: To determine the frequency of identification of the triradiate fold during colonoscopy and evaluate its reliability as a marker of caecal intubation.

METHODS: One hundred consecutive patients undergoing colonoscopy in a tertiary hospital colorectal unit from May to September 2013 were studied. Video documentation of the caecum was recorded and shown to consultant colorectal surgeons on the unit. Each reviewer was asked through a series of questions to independently identify the triradiate fold. The main outcome was the frequency of visualisation of the triradiate fold in the caecum.

RESULTS: The triradiate fold was seen on average in 18% of cases, but inter-observer agreement was poor. There were only four patients (4%) in which all reviewers agreed on the presence of a triradiate fold. In patients who had undergone previous appendicectomy, the appendiceal orifice was less frequently seen compared with patients who had not undergone appendicectomy.

CONCLUSION: The triradiate fold is infrequently seen during colonoscopy and is therefore an unreliable landmark of caecal intubation.

Accurate assessment of the extent of colonoscopy relies on the identification and cognitive integration of several caecal landmarks to the trained endoscopist. These landmarks may include the ileocaecal valve, ileal mucosa, appendiceal orifice and caecal folds.

Of these many landmarks, the ileocaecal valve has been shown to be the single most reliable landmark, being viewed 98% of the time when caecal intubation had been confirmed with fluoroscopy[1]. The same study found the appendiceal orifice to be the second most reliable landmark. Others contend that the most accurate method to ensure caecal intubation is to enter the terminal ileum and confirm with a biopsy of ileal mucosa[2].

Various names and descriptions have been given to the folds in the caecum. The converging folds have been named the triradiate fold, Mercedes sign, crows-foot and caecal strap fold[2-5]. The term triradiate fold is used in many computerized colonoscopy databases and also by the Australian National Bowel cancer screening project as a landmark for caecal intubation with colonoscopy[6]. Last’s Anatomy textbook describes the three taenia of the colon converging at the base of the appendix[7]. However in practice, a clearly defined triradiate fold is not always seen at colonoscopy. We undertook this study to document how frequently a triradiate fold is seen at the caecal pole during colonoscopy.

This study was approved by the Melbourne Health Human Research Ethics Committee at the Royal Melbourne Hospital, Parkville, Australia. All consecutive patients undergoing colonoscopy from May to September 2013 either performed or supervised by two colorectal fellows were included in the study. Exclusion criteria were previous caecal resection, inadequate bowel preparation, technical issues with recording equipment and incomplete colonoscopies.

All colonoscopies were performed with an Olympus colonoscope. Once caecal intubation was reached, a short video was recorded on the Olympus Endobase^® program. The caecal pole was thoroughly irrigated to adequately display the caecum. Caecal intubation was confirmed on collective visualisation of the ileocaecal valve, blind-ending caecal pole and appendiceal orifice when present. It was not routine in our institution to perform ileal intubation as a marker of complete colonoscopy unless clinically indicated. The videos were recorded in a standard fashion to give a panorama of the caecal pole including established landmarks of the appendiceal orifice and ileocaecal valve, and then zoomed in on the appendiceal orifice and surrounding folds.

All 100 videos were then edited using Corel Video Studio Pro^® to delete unnecessary footage. The videos submitted for analysis included continuous footage so that each caecum was easily identifiable. The final length of edited videos ranged from 3 to 26 s. The shortest videos were those where all features of the caecum were very easily seen.

Each video was then shown to six consultant surgeons on the unit who then individually evaluated them. Prior to evaluation, a photograph of what we considered to be a triradiate fold was shown to all surgeons (Figure 1). A photograph showing an appendiceal orifice with no triradiate fold was also shown (Figure 2). For each video the following questions were asked:

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Photograph of a triradiate fold converging on the appendiceal orifice.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Photograph of an appendiceal orifice with no converging triradiate fold.

Are you satisfied that this is a video of the caecum?

Can you identify the appendiceal orifice?

Is there a triradiate fold at the appendix orifice?

If the reviewer was not satisfied that the caecum was represented in the video, the remaining two questions were obsolete and not answered. Similarly if the appendiceal orifice could not be identified, no judgment could be made on the presence of a triradiate fold.

One hundred and thirty-four consecutive colonoscopies were either performed or supervised by two fellows. Of these 134 patients, 34 were excluded, leaving 100 colonoscopies for analysis. Of the patients excluded, 11 had a previous caecal resection, 14 had inadequate bowel preparation, 6 were incomplete colonoscopies and 3 had technical problems with the recording equipment. Of the 100 included patients, there were 43 males and 57 females with a mean age of 61 ± 15.8 years. The indications for their colonoscopies are given in Table 1.

A Conger’s Kappa coefficient for multiple raters was calculated to assess agreement between the answers to each of the three questions. This showed a strong agreement for each of the three questions, with coefficients of 0.93 for identification of the caecum, 0.79 for identification of the appendix and 0.81 for identification of the triradiate fold.

On average, the reviewers were satisfied the video depicted the caecum in 97% of cases. In those cases where the caecum was identified, the appendiceal orifice could be seen in 83%. In those cases where the appendix orifice was identified, a triradiate fold was seen in 18% (Table 2).

The Kappa coefficient for the triradiate fold was high due to strong agreement among observers about the non-visualization of the triradiate fold. In the smaller percentage of cases where the triradiate fold was identified, there was poor agreement among reviewers. There were only four patients in which all reviewers agreed on the presence of a triradiate fold. The appendiceal orifice was seen in 38 of 54 videos (70.4%) where the patient had undergone previous appendicectomy. When the patient had not undergone an appendicectomy the appendiceal orifice was seen in 444 of 527 videos (84.3%, P = 0.014).

Accurate identification of the caecum at colonoscopy relies heavily on visualising certain landmarks. The caecal folds have been previously suggested to be an unreliable landmark[8]. In this study we have demonstrated that a triradiate caecal fold is only seen 18% of the time when averaged across all observers. In cases where it was identified, there was poor inter-observer agreement.

Anatomically, the triradiate fold must be centred on the appendiceal orifice. Our study showed that the appendiceal orifice was less frequently seen in patients who had undergone previous appendicectomy. Elsewhere in the colon the intersection of haustral folds and a taenia coli may create a triradiate appearance that could be confused with the triradiate fold of the caecum.

This study is limited by its subjective design. The six consultants who analysed each video were aware of the study hypothesis, which may have influenced their response to the questions. However, they were not aware that all videos were of the caecum. Variables that may influence the identification of the caecal folds include the amount of insufflation. A large amount of inflation of the caecum may flatten the caecal folds, and thus a triradiate fold that may have been seen in a less distended caecum may have “disappeared” with inflation. Although this theory was not directly tested, we did not observe that inflation had any bearing on the visualization of a triradiate caecal fold through the process of recording, reviewing and editing the videos.

Photo documentation of caecal intubation has been recommended as routine practice although some studies have shown that still photography has poor reliability[9]. Video documentation has previously been shown to be superior to still photography in identifying the caecum to independent observers[3]. Our initial proposal was to use photo rather than video documentation. On review of the first still photographs taken it was our belief that a photograph alone was not sufficient to identify the caecum with certainty. The video however, gave greater detail and provided an accurate depiction of the caecum as demonstrated by the 97% agreement among reviewers. If caecal intubation documentation is to become a marker of quality and successful completion, video documentation appears to be a more reliable method.

This is the first study to look specifically at the triradiate caecal fold as a landmark during colonoscopy. The triradiate fold is an infrequently seen feature of the caecum and as such should not be relied upon to confirm caecal intubation.

We would like to thank the staff of the endoscopy suite at the Royal Melbourne Hospital for their assistance in the technical aspects of video recording.