Over the last few years, there has been increasing attention on surgical procedures to treat haemorrhoids. Several comparative studies have been performed to evaluate the procedures already available to treat second, third, and fourth-degree haemorrhoids, and new surgical techniques, such as, for instance, haemorrhoidectomy with Harmonic scalpel^® [1–3] and Ligasure™[4], doppler guided haemorrhoidal plexus ligation[5, 6] and the stapled haemorrhoidopexy [7–13].

The most recent publications provide data on the efficacy, the results and complications from medium and long-term studies carried out on ample case series [14–20]. Despite the support offered by different authors, none of these techniques proved to reduce pain at such levels to be unanimously chosen [11].

Today, however, the Milligan-Morgan haemorrhoidectomy is still one of the most popular surgical treatments of haemorrhoids

The aim of the present work is to assess postoperative pain, together with other early and late complications, after Milligan-Morgan haemorrhoidectomy as we could observe in our experience before and after performing an internal sphincterotomy.

In 31 cases none of the mentioned symptoms and complications were present, but the patients insisted to undergo surgery.

Before the procedure, all patients underwent proctoscopy.

The operations were carried out under general anaesthesia and infiltration of pudendal nerves with 15 ml bupivacaine with 1:200000 adrenaline. Further 5 ml of the same solution were used to dissect the haemorrhoidal nodules from the internal sphincter. Except 27 patients, who had a considerable rectal mucosal prolapse and were treated with stapled haemorrhoidopexy, all the patients underwent the same procedure, i.e. the Milligan-Morgan haemorrhoidectomy

We removed the haemorrhoidal nodule by performing an upside-down V-shaped incision on the anal dermis, without widening the surgical wound while approaching the sphincters. This was done in order to maintain ample mucous membrane bridges. Possible secondary nodules were removed trough submucosa. Furthermore we performed these technical arrangements:

Since 1995 an internal sphincterotomy of about 1 cm was performed at the end of the procedure, as was the case for 220 patients. We chose permanent surgical sphincterotomy, instead of temporary medical one, because at that time it was the only effective procedure in reducing postoperative sphincter spasm.

The patients were divided into two groups: patients operated on before (group A) and after 1995 (group B). The data concerning the complications of this group of patients (group B) were compared with those regarding the patients operated on before sphincterotomy (group A). No antibiotic prophylaxis has ever been administered.

In particular, we evalued pain using the visual analog scale (VAS). We considered pain as severe when: VAS > 5; patients needed administration of ketorolac three times or more a day; pain did not disappear after the third postoperative day.

Three years after the operation the patients were contacted again by phone to check the recurrence of haemorrhoids and/or other disorders and, where possible, to have a further check-up.

The study received the approval from the University School of Medicine of Palermo Ethics Committee and was in compliance with the Helsinki Declaration.

Of all 699 patients who underwent haemorrhoidectomy, 507 patients (72.5%) did not have any postoperative complication. Table 5 shows the postoperative data concerning 192 patients (27.46%) out of 699.

Pain stands as the most frequent complication (23.03%). These patients had to be administered ketorolac more than three times a day. However, pain tended to disappear spontaneously between the third and the fifth postoperative day. In 19 cases (2.7%) pain persisted for more than ten days and was accompanied by the onset of an anal fissure. The number of patients who suffered from postoperative pain decreased significantly when performing internal sphincterotomy, going from 28.8% down to 10.45% (χ² : 10,880; p = 0,0001); 95% Confidence Interval (CI) 24.7 to 28.9 (group A) and 10.17 to 10.72 (group B). We did not register any clinical impairment of continence in the two groups of patients.

In 51 cases (7.29%) urinary retention was registered (in 37 cases this complication was associated with pain or oedema or fever). The problem was solved by applying a catheter. The sphincterotomy did not reduce such complication significantly.

Six cases of bleeding (0.85%) were registered: four early cases, between the first and the second postoperative day, and two late cases. Despite the various efforts made, haemostasis was achieved by performing a transfixed stitch. The two cases of late haemorrhage showed more serious consequences. One patient had a serious anaemia with haemoglobinemia levels of 7.5 g/dl. The other patient, on the tenth day, did not show serious anaemia thanks to a further early procedure. However, he had to be transferred to intensive care, due to the onset of aspiration pneumonia.

One month after the operation 697 patients had a check-up: in 221 of these patients (31.61%) healing had not been achieved yet, whereas 5 cases (0.71%) had already been affected by stenosis. Finally, 13 patients (1.87%) complained that they sometimes were not able to control gas (8 cases) or also liquid faeces. There was not any statistically difference as regard the impairment of continence in the two groups of patients (table 4).

Six months after the operation 694 patients were evalued: 477 had the operation before 1995, whereas the remaining 217 patients had the operation after 1995. Twenty-three patients (3.31%) were diagnosed as having a fissure on one of the haemorrhoidectomy beds. The internal sphincterotomy did not reduce significantly this complication, which affected 2.59% of cases. It is worth emphasising that 21 patients out of the 23 patients who were affected by a fissure had had pain as an immediate postoperative complication. Three other patients were affected by stenosis, whereas patients suffering from incontinence decreased to nine (1.30%) and only occasionally had gas incontinence.

After three years, the results were studied on 532 patients out of the 699 patients who had been operated and answered the questionnaire. Thirteen patients (2.45%) had a recurrence and 47 patients (8.83%) reported that they occasionally experienced bleeding, though there was no evidence of haemorrhoidal recurrence. In 38 patients skin tags were found.

Drawing on our experience, we evalued the complications arising after surgical procedures to treat haemorrhoids and in particular we wanted to assess postoperative pain and other complications.

Among the complications, the most frequent one is surely the onset of severe pain, whose incidence in our data is in line with literature data [10–21] including studies on the use of a stapling device.

In this respect, it is worth considering the importance of performing anal divulsion/internal sphincterotomy to prevent postoperative pain, as already highlighted in several multi-centre clinical trials [22, 23]. Some authors, instead, assess that the internal sphincterotomy does not influence the incidence of postoperative pain [24, 25]. Sometimes pain may persist over time after cicatrisation of wounds. In most of these cases, the cause should be attributed to the formation of an anal fissure on the groove of one of the excised peduncles.

Continence-related disorders include the possible onset of various degrees of incontinence or of costiveness. Incontinence occurring after a procedure of low ligature is rare and transitory, as it resolves by itself within 3-5 days after the operation. Moreover, it is usually not above grade C Browning-Parks[26].

In our experience, the incidence of incontinence was lower than the levels reported by Graviè et al. [17] at check-up after one month (1.86% vs. 8.8%) and after six months (1.29% vs. 8.8%). Incontinence is more frequent after a Whitehead procedure, and it is also for this reason that this technique is not popular and it is indicated for selected patients under the supervision of an expert surgeon [27].

In reviewing 400 patients treated by Montorsi [28], the same author found a proportion of anal fissures amounting to 5%. In such cases, the treatment consists in regularizing the alvus and, in particularly resistant cases, performing an anal divulsion or an internal sphincterotomy.

The cases of urinary retention observed in our study (7.30%) are less than those indicated by Toyonaga T. et al. [29, 30] (21.9%), and they are near the data provided by Chik B. et al. [31] (7.77%) in a study on stapled haemorroidopexy. This complication affects more male subjects, mostly aged between 40 and 60 years.

Postoperative bleeding is a particularly important complication in treating haemorroids due to its frequency, which vary between 0.6% and 10% [31, 32] depending on the study considered. Sometimes bleeding may be alarming, because it may cause anaemia very rapidly in patients.

The causes of such bleeding are not easily explained: in some cases it should be attributed to falling off of an scar due to electrocoagulation, whereas in other cases it is due to the lack of a thrombus, its expulsion or its dissolution, concomitant with the falling or reabsorption of the transfixed stitch.

Haemorrhoidal recurrence stands at around 2-8% [17]. On the other hand, in the past the proportion of recurrence used to be around 14%, as already pointed out by Goligher [33]. However, one system to prevent this from occurring is to perform subcutaneous removal of these nodules.

Long-term outcomes of haemorrhoidectomy procedures include the onset of malformations of the anal canal, with varying degrees of seriousness up to cases of stenosis. In our experience, the incidence of anal stenosis is very low (0.43%). This is probably due to the fact that we avoid widening the residual surgical wound.

As regards long term recurrence (three years), in our data we found an incidence of 2.45%, which is in line with literature data and lower than methods that are considered more efficacious, such as, for instance, stapled haemorroidopexy, which is reported to have a higher incidence[34].

The types and frequency of immediate, medium and long-term complications we could observe following a Milligan-Morgan haemorrhoidectomy are in line or, for some variables, even less than those described in the literature.

The present paper shows how, according to our experience, the internal sphincterotomy: reduces significantly pain only in the first postoperative period, but not in the medium-long term follow up; does not increase the incidence of continence impairment when performed; does not influence the incidence of the other postoperative complications especially as regard medium and long term results.

Our study highlights how results can be improved by implementing some technical arrangements, which expert proctologists are able to carry out. This way, in terms of results and complications, the Milligan-Morgan haemorrhoidectomy is similar to other techniques, which are recently become popular because of the lower degree of pain referred by their advertising.

GD: Full Professor of Surgery at the School of Medicine of the University of Palermo, Director of the Unit of General and Geriatric Surgery at the University Hospital Policlinico "P.Giaccone".

GG: researcher at the School of Medicine of the University of Palermo, resident surgeon at the Department of General, Urgency and Transplant Surgery.

BC: researcher at the School of Medicine of the University of Palermo.

CR: fellow in Digestive and Endoscopic Surgery at the School of Medicine of the University of Palermo.