It is a randomized, parallel-group (1:1:1), double-blinded, single-center prospective study. No changes in methods of the study were allowed after commencement. Flowchart of the study is presented in Fig. 1. This study was performed in Vilnius University Hospital Santara Clinics, Vilnius, Lithuania. It is a large tertiary university hospital with a dedicated outpatient clinic. The trial was approved and registered at the Regional Bioethics Committee of Vilnius, Lithuania, on the 6th of October, 2015, registration number 158200–15–792-322 (Appendix figure 1). All patients signed the informed consent form for participation in the study in addition to the consent form for the operation.
Inclusion and exclusion criteria
Patients with symptomatic 2nd- or 3rd-degree hemorrhoids, in the 1st to 3rd risk group of ASA (American Society of Anesthesiologists), who consented to participate were included in the study. Exclusion criteria were 1st- or 4th-degree of hemorrhoids, pregnancy, patients with other anorectal diseases (fistula, abscess, rectal carcinoma, inflammatory bowel disease, etc.), patients after previous anal operations (except rubber band ligation, which should have occurred more than 3 months before the inclusion into the trial), and 4th or higher risk group of ASA.
Preoperative evaluation
The detailed physical and anorectal examination was performed with anoscopy and rigid proctoscopy in all cases, as well as colonoscopy if indicated. All patients filled a dedicated symptom questionnaire, which included questions on the intensity and frequency of hemorrhoidal prolapse, bleeding, itching, pain, and other symptoms. Every patient completed Wexner incontinence score [11], SF-36, and fecal incontinence quality of life (FIQOL) [12] questionnaires before surgery.
Randomization, blinding, and concealment
The patients were randomized into three groups. The randomization sequence was computer-generated before the start of the trial. Every consecutive case history was assigned a randomization number (1, 2, or 3 for each treatment modality). It was written and sealed within the envelope and remained unknown neither to the patient nor to the treating physician, to avoid selection bias. In the operating room, after induction of anesthesia, operating room junior staff was asked to unseal the envelope, and the intervention was performed according to the procedure assigned. Pre- and postoperative patient management was as close to identical as we could make it in all three groups. The patient remained unaware of the procedure performed until the end of the study, 1 year after the operation. The case notes and discharge summary of the patient contained the note, saying that “the patient is included in the study of hemorrhoids, patient’s number is X.” This number was within the locked and coded database. The surgeon, who evaluated the result of the treatment remained blinded to the procedure performed. The patients were followed up by different surgeons (EP, VJ, and KS) than the ones performing the operation. They had access to the patient notes but not to the coded database and were not able to know which procedure had been performed in a patient. In emergency situations, unblinding of the patient and treating physicians was possible but was not required or performed in any of the patients.
Operative procedure
Patients were started on lactulose the day before the operation, which was continued after the operation to have regular bowel movements.
Preoperative intravenous antibiotic prophylaxis was given according to the hospital protocol, which was 1 g of cephazolin (2 g if a patient was over 80 kg of weight), 240 mg of gentamycin, and 500 mg of metronidazole (ciprofloxacin could be used if a patient was allergic to cephalosporins).
Each surgeon performing operative procedures (TP, DD, and SM) had personal experience of at least 50 operations of each modality. A 1-h seminar was conducted between all surgeons before the start of the trial to unify the technique of operative procedures.
All patients were photographed after induction of anesthesia before the start of the procedure and immediately after the procedure.
LHP was performed using a Ceralas diode laser of 1470 nm wavelength (Biolitec). Disposable LHP kit (Biolitec) was used, which contains sharp-tipped laser fiber and anoscope. Perianal skin immediately aboral to hemorrhoid was penetrated using needle tip cautery (Fig. 2). Laser fiber was introduced into the opening until the level of hemorrhoidal pedicle and coagulation was activated. A total of 8 W 3 s pulses with 1-s pulse-pauses were used to coagulate the area of hemorrhoids. A total of 4 mm of hemorrhoidal tissue is coagulated with one such pulse. A total of 250 J was the upper limit of energy delivered per 1 hemorrhoidal quadrant. Smaller hemorrhoids were treated with less energy. The procedure was repeated in the other three quadrants, thus treating all anal circumference.
MP was performed as described by Schurmann JP et al. [8]. Ligations were placed in the area of visible hemorrhoidal tissue. It was started with a single suture ligation at the level of hemorrhoidal pedicle which should incorporate the feeding vessel and continued down until above the dentate line. The continuous suture was tied, thus lifting the prolapsing hemorrhoidal tissue.
Standard EH was performed up to the level of the hemorrhoidal pedicle, with ligation or suture ligation of the pedicle and meticulous hemostasis.
Follow-up
Follow-up was performed by the different surgeons to those performing the operations (EP, VJ, and KS). Each of them had more than 25 years of experience of colorectal and hemorrhoidal surgery. The 1-h seminar was conducted, with them to unify the evaluation of the patients within the study. None of the evaluators knew which operation was performed, and they had no access to coded database of operations performed. Patient documents only mentioned that they were included in the study and study number but not which operation had been performed.
Each patient was followed up at 1 and 6 weeks and after 1 year after the operation. Perianal examination with photographic documentation was performed during all visits. Anoscopy was performed during visits at 6 weeks and 1 year.
Each patient was asked to fill in the diary every day of the first postoperative week and present it at the first visit after 1 week. Symptom questionnaires were filled in during visits at 1 and 6 weeks and 1 year. Wexner incontinence score, FIQOL, and SF-36 were filled in by the patient during the visit at 1 year.
Outcomes
The primary outcome of the study was the recurrence of rectal bleeding and prolapse at 1 year after the operation requiring any kind of medical attention or treatment (visit to the doctor or pharmacy, medical, invasive, or surgical treatment).
Secondary outcomes of the study were time to return to work or regular activity, intensity, and duration of perianal pain after the operation (in days), Wexner fecal incontinence score at 1-year visit, QOL based on SF-36 questionnaire and FIQOL at 1 year, and evaluation of the operation by the patient on visual analog scale from 1 to 10 at 1-year visit.
No change in outcomes or outcome evaluation was allowed after the commencement of the study.
Statistical methods
The sample size required for the study was calculated using the flexible statistical power analysis program for the social, behavioral, and biomedical sciences, G*Power V 3.1.9.2.
A sample size of 40 patients in each randomized group provides 84% power to detect an effect size of 0.30 in the recurrence of symptoms at 1 year across the randomized groups, with an alpha of 0.05.
We used Shapiro-Wilk and Kolmogorov-Smirnov (K-S) tests for the verification of the normality of variables. A statistically significant relationship between the related variables was determined using a number of criteria. For the variables that satisfied the condition of normality, we used parametric ANOVA criteria, which is also called the Fisher analysis of variance, and it is the extension of the t and z tests. For the variables that did not satisfy the condition of normality, we used nonparametric criteria based on the χ2 criterion for the interval and categorical variables, i.e., the Kruskal-Wallis test, which is the nonparametric test equivalent to the one-way ANOVA, and an extension of the Mann-Whitney U test to allow the comparison of more than two independent groups (in our case, we have three independent patient groups divided by the type of surgery). When the data were described in a four-digit (2 × 2) frequency tables and when at least one expected value was less than 5, Fisher’s exact test was additionally calculated. We used the Pearson chi-squared criterion to compare two independent groups. We used the Wilcoxon test to compare two dependent groups. We used the marginal homogeneity test to compare the categorical data. We used Student’s t test to compare the data that satisfied the condition of normality. The degree of linear dependencies of variables (correlation coefficients) of the Spearman or Kendall τ-b was calculated for interval variables, when the normality was not satisfied and for the rank variables.
The dependency between variables can be considered statistically insignificant when the two-sided p value (exact sig. (2-sided), approx sig.) of all criteria in this category is higher or equal than the established significance level of 0.05 (p ≥ 0.05) and statistically significant when p < 0.05. The confidence interval (CI) was calculated for a 95% confidence level.
Statistical analysis was performed using software packages, i.e., R statistical software package V 3.5.3 (2019-03-11) (©The R Foundation for Statistical computing), R studio V 1.1.463–© 2009–2018 R studio Inc., IBM SPSS Statistics V.23.