Introduction Successful treatment of an anal fistula requires correct identification of the internal opening and accurate delineation of the course of the fistula in relation to the anal sphincters. Goodsall’s rule states that a fistula with the external opening anterior to an imaginary transverse line across the anus has its internal opening at the same radial position and for an external opening posterior to this line, the internal opening is in the midline posteriorly with a horse-shoe track. The aim of this study was to determine the accuracy of Goodsall’s rule and the relationship between the distance of the external opening from the anal verge and the configuration of the fistula.

Method We studied a consecutive series of 103 patients (20.4% female, 79.6% male) undergoing their first operation for cryptoglandular anal fistula between 2006 and 2014, by a single colorectal surgeon. The site of the internal and external openings were recorded by reference to hours on a clock (lithotomy position) and the radial distance of the external opening from the anal verge was measured with a thread. A horseshoe fistula was defined as one where the internal and external opening were separated by 2 or more hours. Anterior fistulas lay between 9 and 3 o’clock (58/103; 56.3%), posterior between 4 and 8 o’clock (45/103; 43.7%).

Results The diagnostic accuracy of Goodsall’s rule was only 66.9%. Goodsall’s rule was correct in 25/58 (43.1%) anterior fistulas and 30/45 (66.8%) of posterior fistulas. Of the transsphincteric fistulas, 31/51 (60.8%) conformed to Goodsall’s rule.

The distance between the external opening and the anal verge influenced the accuracy of Goodsall’s rule, being true in 92.9% of those <15 mm, 80.9% of those between 16–25 mm, 70% of those between 26–35 mm and 47.4% of those >36 mm. Distance of the external opening from the anal verge influenced the likelihood of the fistula having a horseshoe course: <15 mm (0%),16–25 mm (9.5%) 26–35 mm (33.3%), >36 mm (57.9%). For fistulas with an external opening in the anterior half of the anus, 4/5 (80%) ≤20 mm from the anal verge had a radial course compared to 29/53 (54.7%) where the distance from the verge was ≥21 mm. For those with an external opening in the posterior half, 18/18 (100%) ≤20 mm from the anal verge had a radial course compared to 18/27 (66.7%) where the distance from the verge was ≥21 mm.

Conclusion Goodsall’s rule is of limited use in predicting the site of the internal opening of a fistula. The closer the external opening is to the anal verge, the more likely a fistula conforms to the rule. The further a fistula is away from the anal verge, the more likely it follows a horseshoe course and less likely to conform to Goodsall’s rule.

Disclosure of interest None Declared.