Preventing hemorrhoids during pregnancy: a multicenter, randomized clinical trial

A randomized, single-blind, multicenter trial was conducted between June 2016 and June 2019 in three different clinical centers (Vilnius University Hospital Santaros Klinikos; Vilnius City Clinical Hospital and Vilnius Maternity Hospital). Women in early pregnancy (less than 12 weeks of gestation) were informed about the study. If they showed interest in participating, they were screened for eligibility. The study was approved by the Vilnius Regional Bioethics Committee, Vilnius, Lithuania on the 10th of May 2016, registration number 158200–16-843–357 (registration certificate provided as a supplementary document).

Women with early viable pregnancy (less than 12 weeks of gestation) at the age between 18–45 years and who gave written consent were included in this trial. All other women who did not fulfill all the inclusion criteria were not eligible to participate in the study.

Patients were randomly allocated into two groups in a ratio of 1:1. A computer-based randomization sequence was generated, transferred and sealed into individual envelopes. Once a patient gave written consent to participate a clinician unsealed an envelope in sequence and the patient was allocated into one of the two groups.

The study intervention was designed by using the following guidelines and recommendations [11, 12]. During the first visit, each woman in the intervention group received a structured 30-min personal consultation with written instructions of dietary and behavioral recommendations. Women were advised to eat at regular time intervals; Consume at least 1.5 L of fluid, avoid food, that causes constipation; Consume a tablespoon of bran and 2–5 prunes daily; Consume around 300 g of fruits, 500 g of vegetables and 30 g of nuts daily; Exercise and/or walk daily 30–60 min, 3–5 times per week. Furthermore, there were specific recommendations for defecation: not to ignore the urge to defecate; Spend less than 3 min on the commode; Attempt to defecate 30–40 min after eating and in the mornings; Washing after bowel movement. Full detailed intervention recommendations are provided as supplementary material. Women in the control group received standardized, nationally approved physical activity and dietary recommendations for pregnant women.

Each participant had a total of three study visits (Fig. 1.). The first visit took place during the first trimester of pregnancy (< 12 weeks of gestation), second visit during the second trimester of pregnancy (18–20 weeks of gestation) and the third visit was carried out upon discharge from the obstetrics unit 2–3 days after childbirth. The First visit coincided with the study enrollment, during which a detailed questionnaire consisting of socioeconomic factors, physical activity, anthropometric data (weight and height), obstetric history, perianal symptoms in previous pregnancies and the presence of chronic health conditions was filled out. The proctologic questionnaire was filled out and the physical examination was performed during each visit. On the first and second visits, detailed information about dietary habits, physical activity, alcohol and tobacco consumption was collected. Pregnancy outcomes and neonatal data were gathered from medical records during the third visit.

The primary outcome of this study was the rate of hemorrhoids at the time of discharge from the obstetrics unit. Outcome was assessed by a gynecologist who remained blinded to the patients’ allocation group. To keep the evaluation of the outcome measure consistent, all participating gynecologists underwent a 4-h seminar conducted by the same expert proctologist on how to assess the presence of hemorrhoids using a standardized methodology.

Secondary outcomes of the study were the safety of the intervention, measured by the rate of miscarriages in both groups, and the possible risk factors to develop hemorrhoids during pregnancy.

The sample size was calculated using the G*Power software. We presumed the baseline risk of hemorrhoids during pregnancy of 35% from our previous experience and we predicted that the intervention would reduce the risk to 17% [3]. Based on the statistical power of 80% and a level of significance set at 5% we calculated the total sample size to be 206 patients. The sample size was increased by 30% to 260 patients to account for loss to follow-up and miscarriages. As the intervention: control ratio was 1:1, each arm of the study consisted of 130 participants.

Categorical variables were compared by the Chi-Squared or Fisher exact tests, as appropriate. Continuous variables were compared using either the Student's t-test or the Mann–Whitney U test. Binary logistic regression model was used to identify independent predictors of hemorrhoids after delivery. This analysis was performed on factors with a p-value < 0.10 in univariate analysis. Statistical analysis was performed using IBM SPSS 23.0 and GraphPad Prism 9 software. A P-value of less than 0.05 was considered significant for all tests.

Our study is the first randomized controlled trial to prove that a counselling intervention, aimed to modify dietary and behavioral habits, can significantly reduce the rate of hemorrhoids in pregnancy. Both ITT and PP analyses showed that this intervention managed to decrease the hemorrhoids rate by about half (ITT (RR 0.38; 95% CI 0.24–0.59; p < 0.001) PP (RR 0.42; 95% CI 0.27–0.64; p < 0.001). The incidence of hemorrhoids after giving birth in the control group (ITT-40.8%; PP-47.7%) was in line with the reported rate of 40.7% observed by Poskus et al. in a similar population [3].

Pregnant women are a very vulnerable population therefore the safety of intervention is crucial. We chose to analyze the miscarriage rate in order to prove that our intervention did not prompt unfavorable pregnancy outcomes. The miscarriage rate in both groups did not differ significantly, furthermore, the patients did not report any additional side-effects that could be attributed to the effects of an intervention.

We identified that the history of perianal disease and newborn height were independent risk factors to develop hemorrhoids after delivery on multivariate analysis. Intervention was the only protective factor significantly reducing the likelihood of hemorrhoids. Our findings are somewhat similar to those reported by Ferdinande et al. and Poskus et al. as they also determined that a previous history of perianal disease is highly associated with the increase of hemorrhoids rate during pregnancy [3, 5]. However, we did not find that constipation before the first trimester would be associated with hemorrhoids after delivery.

Intervention applied in our study contained dietary and behavioral habit changes which are also recommended for conservative hemorrhoids treatment in non-pregnant patients [13, 14]. Our intervention is safe, cost-effective and does not require any additional training for the medical personnel or patient. This allows to apply this intervention widely without limiting it to specialised treatment centers. In our opinion, counselling could be a task for primary care as modifying risk factors and changing behaviour when planning a pregnancy may yield even better results.

The main limitation of this study is the number of patients who were lost to follow-up (14 (10.8%) women in the control and 28 (21.5%) in the intervention group). This, perhaps, reflected the population (working age) and the condition (considered by some to be a sensitive subject). Having this in mind, we performed both ITT and PP analyses, which gave us identical results, showing that these lost to follow-up patients did not skew our findings. Furthermore, the study’s initial sample size was increased by 30% to account for loss to follow-up and miscarriages thus the study still had sufficient statistical precision to detect differences between groups. Small differences between study groups may have been missed due to the low miscarriage rates, however, miscarriage rate is the most acknowledged outcome when evaluating the safety of various interventions during pregnancy. Our study design was pragmatic and we did not strictly control whether the patient complied with the intervention’s recommendations. However, as both trial arms were balanced on the baseline characteristic and were equal on pregnancy outcomes, we would draw a conclusion that the intervention had the main influence on the reduced rate of hemorrhoids after delivery.