Abstract
Background
Incisional hernias with apertures measuring less than 7 cm can generally be treated adequately with the laparoscopic intraperitoneal onlay-mesh (IPOM) technique. The wearing of an abdominal binder after surgery is often recommended in order to promote wound healing and prevent recurrent herniation. We carried out a multicenter, randomized pilot trial to evaluate the utility of abdominal binders.
Methods
The trial was conducted from May 2019 to December 2020. Persons with a laparoscopic IPOM procedure for treatment of an incisional hernia were included in the trial and randomized preoperatively (1:1). The patients in the abdominal binder group wore an abdominal binder during the day for 14 days after surgery, while those in the control group wore no binder. The primary endpoint was pain at rest on postoperative days 1, 2, and 14, as measured on a visual analog scale. The secondary endpoints were overall subjective well-being, the rates of wound infection, recurrence, and complications, mobility, and the rate and size of postoperative seromas (on postoperative days 1, 2, and 14).
Results
Forty patients were included. Three were excluded because of conversion to an open surgical technique. The biometric and perioperative data of the abdominal binder group (n = 18) and the control group (n = 19) did not differ to any statistically significant extent. The patients in the binder group had significantly less postoperative pain (F [dfn, dfd]) 4.44, 95% confidence interval [1; 35]; p = 0.042).The patients in the binder group also had better overall subjective well-being and a higher rate of postoperative seroma formation, but these differences did not reach statistical significance. There was less limitation of mobility than in the control group; however, this difference also did not attain statistical significance.
Conclusion
An abdominal binder may reduce pain after incisional hernia repair with the IPOM technique. The postoperative use of analgesic medication was not measured.
Incisional hernias (IH) occur in 9–20% of patients following laparotomies (1). Advanced age, diabetes mellitus, obesity, coronary heart disease, and nicotine abuse are among the factors that favor the occurrence of an IH (2– 5).
An IH may lead to pain, discomfort, restrictive ventilation disorders, and bowel incarcerations. In up to 12% of these cases emergency repair is needed (2– 4, 6, 7). A number of surgical procedures that yield adequate results with regard to rates of wound infection and recurrence have been described in the literature. These techniques vary in their operative approach (laparoscopy versus open surgery) and the positioning of the synthetic mesh (inlay vs. onlay vs. sublay vs. intraperitoneal onlay-mesh [IPOM]) (5, 8, 9). The current clinical guidelines recommend mesh placement if the aperture of the hernia gap measures more than 1 cm (2– 4, 6, 7, 10). Wearing of an abdominal binder (AB) is often advised to promote wound healing and reduce the rate of recurrence. However, the level of evidence for these recommendations is low (11– 13).
Some authors have suggested that wearing an AB reduces pain after IH repair (11, 12). However, the relevant literature also includes discussion of possible negative effects (14, 15). To throw more light on the benefits of AB use, we conducted the multicentric randomized pilot study described here.
Methods
Patients who underwent IH repair by means of the IPOM technique at the Department of General, Visceral, and Cancer Surgery at Helios Hospital Berlin-Buch or the Department of General, Abdominal, Vascular, and Transplant Surgery at the University Hospital, Magdeburg between May 2019 and December 2020 were enrolled in the ABIHR-I trial. The patients were randomized into two groups (allocation ratio 1:1): The AB group wore an AB (Verba®, Paul Hartmann AG, Heidenheim) during the day for a period of 2 weeks; the control group did not wear an AB. All endpoints were measured by non-blinded surgeons of the participating departments.
Endpoints
The primary endpoint was pain at rest measured using the visual analog scale (VAS) on postoperative days 1, 2, and 14. The secondary endpoints assessed on postoperative days 1, 2, and 14 were as follows: general well-being, extent of mobility limitation due to surgery (VAS), the rate (yes/no) and size (cm3) of seromas formation as assessed by ultrasound, and rates of recurrence, wound infection, and complications within 14 days of surgery (according to the Clavien–Dindo classification).
Surgical procedures
All patients underwent IH repair in IPOM technique. Prior to surgery, single-shot antibiosis with cefuroxime was administered. First, a subcostal incision (Palmer´s point) was made in the left abdominal wall to insert the optical trocar after gas insufflation via a Veress needle. We then placed two 5-mm trocars along the semiclavicular line under visual guidance. After adhesiolysis, the contents of the hernial sac were repositioned and the hernia was closed with a non-resorbable suture using a Reverdin needle. A synthetic mesh was inserted into the abdominal cavity to cover the hernia with at least 5 cm overlap. Absorbable tacks were used to fix the mesh in place. All patients received ibuprofen 2400 mg/d in combination with pantozole 40 mg or metamizole 4 g/d. Patient discharge on postoperative day 2 was intended.
Inclusion criteria
Only persons who were cognitively capable of consenting to take part in the study and underwent elective IH repair using the IPOM technique were included.
Exclusion criteria
Pregnant women and patients known to be infected with human immunodeficiency virus (HIV) were excluded.
Statistical methods
The statistical analysis was carried out in R Version 4.0.2 (16) with lme4 (17) and sjstats (18) and GraphPad Prism Version 8.0. Differences between the AB and control groups with regard to the primary and secondary endpoints with continuous variables were examined using linear models with mixed effects. Group (AB, control group) and time (postoperative days 1, 2, and 14) were included as fixed factors and proband as a random factor. The statistical significance was set at p < 0.05. To take account of the fact that the post-hoc analysis involves comparison of three different points in time, Holm–Šídák correction was carried out. Mixed models can use maximum likelihood methods to include missing values in the dependent variables (19, 20). Notwithstanding, to ensure that attrition on the 2nd and/or 14th postoperative day does not influence our result, a sensitivity analysis with imputed values was also carried out. Details on the descriptive statistics, the model with mixed effects, and the sensitivity analysis can be found in the eMethods.
Sample size calculation
No sample size calculation took place due to lack of reference values. A total of 20 individuals for each study arm was considered an appropriate sample size for a pilot study.
The COVID-19 pandemic resulted in significantly reduced patient recruitment. To ensure the rapid planning and execution of a follow-up study with a power-calculated number of cases, the latter was reduced by amendment from 60 to 40 (Eth-05/19, 5 January 2021).
Randomization
Randomization lists were used and pseudo-random numbers were generated in R (ver. 4.0.2). The principal investigator performed the randomization before surgery.
Results
A total of 40 persons were enrolled in this study. Three individuals were excluded after the intraoperative decision to proceed with open surgery (Figure 1 [21]). In these cases, laparoscopic closure or reduction of a hernia repair was not feasible due to the size of the hernia (> 7 cm). All other patients, including those with incomplete follow-up, were analyzed.
Figure 1.
Flow chart of the ABIHR-I pilot study
All of the patients included gave their consent to take part in the study and agreed to ultrasound examination of the abdomen on postoperative days 1, 2, and 14. Twelve patients (66.7%) from the AB group and 17 patients (89.5%) from the control group underwent an (ultrasound) assessment on postoperative day 14. In three cases no sonography of the wound was performed (figure 1).
Univariate analyses of the baseline characteristics and the perioperative data are summarized in Tables 1, 2, and eTable 1. No statistically significant differences were found between the two groups (p > 0.05, corrected for multiple comparisons with the false discovery rate). Because of our moderate sample size, the p-values were considered and interpreted together with the effect size measure. The differences between the AB and control groups show little or no effect for either the baseline characteristics or the perioperative data. The greatest differences between the two groups are in age and ASA stage (table 1).
Table 1. Univariate analysis of baseline characteristics.
| Variable |
AB group
(n = 18) |
Control group
(n = 19) |
Statistics | |||
| t/χ 2 (df) | Effect size | p-value | ||||
| Age | Years | 61.7 (12.0) | 56.2 (15.2) | 1.22 (35) | 0.402 | 0.231 |
| Sex | Male | 11 (61.1%) | 10 (52.6%) | 0.27 (1) | –0.086 | 0.603 |
| ASA stage | I | 1 (5.6%) | 7 (36.8%) | |||
| II | 11 (61.1%) | 8 (42.1%) | ||||
| III | 6 (33.3%) | 3 (15.8%) | ||||
| IV–V | 0 (0.0%) | 1 (5.3%) | 6.95 (3) | 0.433 | 0.073 | |
| Diabetes mellitus | 4 (22.2%) | 3 (15.8%) | 0.25 (1) | 0.082 | 0.618 | |
| Coronary heart disease | 2 (11.1%) | 3 (15.8%) | 0.17 (1) | –0.068 | 0.677 | |
| Oral anticoagulation | 4 (22.2%) | 6 (31.6%) | 0.41 (1) | –0.105 | 0.522 | |
| Immunosuppression | 1 (5.6%) | 0 (0.0%) | 1.09 (1) | 0.171 | 0.298 | |
| BMI | kg/m2 | 30.9 (6.53) | 30.8 (7.50) | 0.05 (35) | 0.014 | 0.960 |
AB, abdominal binder; ASA, American Society of Anesthesiologists physical status classification; BMI, body mass index
Data are presented as mean (standard deviation) for continuous variables or total number (percentages) for categorical variables.
Effect sizes are presented as Cohen’s d, phi, and Cramer’s V for independent t-tests, 2 × 2 contingency tables, and 2 × > 2 contingency tables, respectively.
Table 2. Univariate analysis of perioperative data.
| Variable | AB group (n = 18) | Control group (n = 19) | p value | |
| Hernia gap width | W1 | 155 (83.3%) | 17 (89.5%) | 0.350 |
| W2 | 3 (16.7%) | 1 (5.3%) | ||
| W3 | 0 (0.0%) | 1 (5.3%) | ||
| Primary IH | Yes | 18 (100%) | 16 (84.2%) | 0.079 |
| Operating time | Minutes | 71.6 (45.1) | 62.4 (31.4) | 0.478 |
| Hernia gap closure | Yes | 17 (94.4%) | 18 (94%) | 0.969 |
| LOS | Days | 2.06 (0.42) | 2.16 (0.50) | 0.505 |
| Compliance+ | 12/12 (100%) | 16/16 (100%) | ||
| Surgeon’s experience | < 5 years | 0 (0.0%) | 2 (10.5%) | 0.417 |
| 5−10 years | 11 (61.1%) | 9 (47.4%) | ||
| > 10 years | 5 (27.8%) | 7 (36.8%) | ||
| > 20 years | 2 (11.1%) | 1 (5.3%) | ||
AB abdominal binder; IH incisional hernia; LOS length of hospital stay
Data are presented as mean (standard deviation) for continuous variables or total number (percentages) for categorical variables. + Compliance in terms of (not) wearing an AB for two weeks, queried on postoperative day 14: W1 < 4 cm, W2 ≥ 4 – 10 cm, W3 ≥ 10 cm
eTable 1. European Hernia Society (EHS) classification of incisional hernias.
| Hernia classification | Abdominal binder group (n = 18) | Control group (n = 19) | |
| M1 | W1 | 1 (5.56%) | 0 (0.00%) |
| M1–2 | W1 | 0 (0.00%) | 1 (5.26%) |
| M1–2 | W2 | 1 (5.56%) | 0 (0.00 %) |
| M1–5 | W3 | 0 (0.00%) | 1 (5.26%) |
| M2 | W1 | 4 (22.2%) | 3 (15.8%) |
| M2–3 | W1 | 3 (16.7%) | 2 (10.5%) |
| rM2–4 | W1 | 0 (0.00%) | 1 (5.26%) |
| M2–4 | W1 | 0 (0.00%) | 1 (5.26%) |
| M2–4 | W2 | 1 (5.56%) | 0 (0.00%) |
| M3 | W1 | 5 (27.8%) | 6 (31.6%) |
| rM3 | W1 | 0 (0.00%) | 1 (5.26%) |
| M3 | W2 | 1 (5.56%) | 0 (0.00%) |
| M3–5 | W1 | 1 (5.56%) | 0 (0.00%) |
| M4 | W1 | 0 (0.00%) | 1 (5.26%) |
| L3 | W1 | 1 (5.56%) | 1 (5.26%) |
| rL4 + M5 | W2 | 0 (0.00%) | 1 (5.26%) |
Data are presented as total number (percentages).
The results of the mixed models of the primary and secondary endpoints are summarized in Table 3 and shown in Figure 2. Compared with the control group, the patients in the AB group had on average significantly less pain at all three postoperative visits. The difference in pain perception between the two groups shows a moderate effect. Adjusting for differences in the ASA stage of the two groups does not change the significance and strength of this effect (p = 0.033; ηp2 = 0.06). Moreover, mobility improved significantly from postoperative days 1 and 2 to postoperative day 14 in both groups. For the remaining secondary endpoints, including general well-being and seroma size, there were no significant group and/or time interactions or main effects. Furthermore, both the differences between the AB group and the control group and the differences among the three postoperative visits for these secondary endpoints show little or no effect. The descriptive statistics and the post-hoc comparison of the two groups on the three postoperative days are summarized in Table 3 and eTable 2.
Table 3. Results of the mixed model for primary endpoint and secondary endpoints with continuous measures.
| Interaction group × time | Main effect of group | Main effect of time | |||||||
| F(dfn. dfd) | ηp2 | p | F(dfn. dfd) | ηp2 | p | F(dfn. dfd) | ηp2 | p | |
| Primary endpoint | |||||||||
| Pain at rest | 0.06 (2.59) | 0.00 | 0.945 | 4.44 (1.35) | 0.06 | 0.042 | 3.43 (1.6; 47.8) | 0.08 | 0.050 |
| Secondary endpoint | |||||||||
| General well-being | 0.18 (2.59) | 0.00 | 0.839 | 3.42 (1.35) | 0.04 | 0.073 | 0.26 (1.8; 53.7) | 0.01 | 0.750 |
| Limited mobility*1 | 0.16 (2.59) | 0.00 | 0.852 | 1.03 (1.35) | 0.01 | 0.316 | 9.71 (1.7; 49.5) | 0.15 | 0.001* |
| Seroma size | 0.92 (2.59) | 0.01 | 0.406 | 0.17 (1.35) | 0.00 | 0.679 | 0.78 (1.1; 29.2) | 0.02 | 0.402 |
Significant p-values are printed in bold.
For the secondary endpoints, p-values marked with an asterisk remain significant after correction for multiple comparisons using false discovery rate (FDR).
*1 Related to performance of incisional hernia repair
Figure 2.
Pain at rest, general well-being, limited mobility and seroma size on postoperative days 1, 2, and 14 (POD 3). AB, Abdominal binder; CG, control group; POD, postoperative day
eTable 2. Descriptive statistics and post-hoc analysis of primary and secondary endpoints.
| AB group | Control group | Post-hoc analysis*2 | |||||||
| n | M (SD) or % | n | M (SD) or % | MDiff | t(df) | p-value | |||
| Primary endpoint | |||||||||
| Pain at rest on | POD 1 | VAS score | 18 | 33.6 (14.6) | 19 | 45.4 (21.2) | −11.8 | 1.98 (32.1) | 0.159 |
| POD 2 | VAS score | 18 | 28.9 (18.1) | 17 | 38.5 (28.5) | −9.59 | 1.18 (26.9) | 0.407 | |
| POD 14 | VAS score | 12 | 21.3 (18.6) | 16 | 32.2 (28.0) | −10.9 | 1.23 (25.7) | 0.407 | |
| Secondary endpoints | |||||||||
| General well-being | POD 1 | VAS score | 18 | 79.4 (15.2) | 19 | 68.9 (20.3) | 10.5 | 1.78 (33.3) | 0.230 |
| POD 2 | VAS score | 18 | 76.7 (17.2) | 17 | 70.6 (24.7) | 6.08 | 0.84 (28.4) | 0.506 | |
| POD 14 | VAS score | 12 | 76.7 (29.9) | 16 | 64.7 (28.8) | 12.0 | 1.07 (23.4) | 0.506 | |
| Limited mobility *1 | POD 1 | VAS score | 18 | 46.9 (18.9) | 19 | 49.5 (23.3) | −2.53 | 0.36 (34.3) | 0.789 |
| POD 2 | VAS score | 18 | 41.7 (25.4) | 17 | 48.8 (24.8) | −7.16 | 0.84 (33.0) | 0.789 | |
| POD 14 | VAS score | 12 | 24.6 (14.1) | 16 | 29.8 (22.5) | −5.17 | 0.75 (25.3) | 0.789 | |
| Wound infection | POD 1 | Yes/no | 18 | 0 (0.00%) | 19 | 0 (0.00%) | |||
| POD 2 | Yes/no | 18 | 0 (0.00%) | 17 | 1 (5.88%) | ||||
| POD 14 | Yes/no | 12 | 0 (0.00%) | 15 | 0 (0.00%) | ||||
| Recurrence | 1. POD | Yes/no | 18 | 0 (0.00%) | 19 | 0 (0.00%) | |||
| 2. POD | Yes/no | 18 | 0 (0.00%) | 17 | 0 (0.00%) | ||||
| 14. POD | Yes/no | 12 | 0 (0.00%) | 15 | 0 (0.00%) | ||||
| Seroma | POD 1 | Yes/no | 18 | 4 (22.2%) | 19 | 6 (31.6%) | |||
| POD 2 | Yes/no | 16 | 8 (50.0%) | 14 | 4 (28.6%) | ||||
| POD 14 | Yes/no | 12 | 5 (41.7%) | 13 | 5 (38.5%) | ||||
| Seroma size | POD 1 | cm3 | 18 | 3.73 (11.9) | 19 | 1.75 (3.18) | 1.98 | 0.68 (19.3) | 0.741 |
| POD 2 | cm3 | 16 | 7.26 (16.4) | 14 | 3.20 (5.95) | 4.06 | 0.92 (19.4) | 0.741 | |
| POD 14 | cm3 | 11 | 1.14 (1.70) | 14 | 6.37 (20.7) | −5.24 | 0.94 (13.2) | 0.741 | |
| CDC up to POD 14 | 0 | 9 | 8 (88.9%) | 16 | 16 (100%) | ||||
| I | 1 (11.1%) | 0 (0.00%) | |||||||
| II | 0 (0.00%) | 0 (0.00%) | |||||||
| III | 0 (0.00%) | 0 (0.00%) | |||||||
| IV–V | 0 (0.00%) | 0 (0.00%) | |||||||
AB abdominal binder; CDC Clavien–Dindo classification; POD postoperative day; VAS, visual analog scale.
Data are presented as mean (standard deviation) for continuous variables and as total number (percentages) for categorical variables.
*1 Related to performance of incisional hernia repair
*2 Post-hoc analyses were conducted only for mixed model analyses of continuous variables. The p-values are corrected for multiple comparison across time using the Holm–Šídák test (NB: in some cases successive adjusted p-values are identical, even when the original p-values were not).
The sensitivity analysis with imputed measured values for postoperative days 2 and 14 confirmed that on average the AB group felt less pain than the control group over all three postoperative visits. The effect was strong and approached statistical significance of <0.05. Statistical significance was reached when the differences in ASA stage between the two groups were taken into account. In both groups, the pain decreased significantly with the passage of time (eMethods, eTable 3).
eTable 3. Mixed analysis of variance (ANOVA) of primary endpoint with multiple imputations (n = 5) for values at postoperative days 2 and 14.
| Interaction group × time | Main effect of group | Main effect of time | |||||||
| F | ηp2 range | p | F | ηp2 range | p | F | ηp2 range | p | |
| Primary endpoint | |||||||||
| Pain at rest | 0.11 | 0.00–0.01 | 0.741 | 3.75 | 0.09–0.11 | 0.053 | 4.26 | 0.10–0.12 | 0.039 |
| Pain at rest controlling for ASA stage | 0.11 | 0.00–0.01 | 0.737 | 3.97 | 0.10–0.11 | 0.046 | 3.87 | 0.10–0.11 | 0.049 |
Significant p-values are printed in bold.
Discussion
The advice on wearing an AB following IH repair is based on a low level of evidence (2– 4, 6, 10). Hence, in 2018 our study group conducted a nationwide survey of surgical departments (n=48) and patients (n=270) to ascertain what postoperative recommendations are made and how they are implemented after repair of an IH (11, 12). Besides a certain period of rest, frequently an AB was prescribed. The mode and duration of binder wearing varied widely. The majority (71.2%) of the patients questioned stated a pain-reducing effect of their AB (12). These findings were confirmed in this randomized pilot study (Table 3 and eTable 2). The postoperative pain level of patients in the AB group was statistically significantly lower than in the control group.
Other positive effects of the AB have been discussed in the literature. Rothman et al. (2014) published a systematic review analyzing the impact of wearing an AB following abdominal surgery. In the absence of statistical significance and in the face of a low level of evidence in the literature, the authors stated that the AB reduces postoperative mental stress and pain (22). In the field of ventral hernia surgery, only one trial has examined the impact of the AB. Christoffersen et al. (2015) conducted a randomized clinical trial (n = 56). All patients underwent laparoscopic repair of an umbilical or epigastric hernia. A total of 28 patients wore an AB for 7 days, and 28 did not use an AB. The primary endpoint was pain on postoperative day 1. In contrast to the results of the present ABIHR-I study, Christoffersen et al. found no statistically significant difference with regard to postoperative pain (23). However, the pain was rated lower than in the control group. Neither Christoffersen et al. nor our study group detected a significant impact of the AB on general well-being. In both trials, individuals who wore an AB had less pronounced impairment of mobility; however, the findings were not statistical significant (Figure 2, eTable 2). Altogether, the findings of our trial and those of the partially comparable randomized trial conducted by Christoffersen et al. seem to confirm the pain-reducing effect of the AB, as stated by questioned patients questioned in earlier surveys (11, 22). The two trials differ, however, in terms of the indications for surgery, the time for which an AB was worn (7 days and nights vs. 2 weeks during the day), the sample size (n = 56 vs. n = 37), and the timing of pain measurement (postoperative days 1, 2, 3, 7, and 30 during physical activity vs. postoperative days 1, 2, and 14 at rest) (23).
In the literature, some authors have postulated that the AB may reduce seroma formation (14, 15, 22). Our clinical trial analyzed a possible impact of the AB on seroma formation. The rate of seroma formation did not differ between the AB group and the control group (Figure 2, Table 3, eTable 2). The cited randomized clinical trial by Christoffersen et al. also detected no significant difference in seroma formation (23). In summary, the research to date in the field of laparoscopic ventral hernia repair has revealed no evidence of a seroma-reducing effect of the AB (23). Further trials are needed to broaden the evidence base.
From our point of view, there are two main reasons for scientific evaluation of the wearing of an AB after surgery. On the one hand, the AB may lead to lower consumption of analgesic medication due to a possible pain-reducing effect. Especially in view of the administration of opioids and the risk of opioid addiction (24), further research into this topic appears urgently necessary. On the other hand, possible negative effects should be identified and prevented. For instance, it has been postulated in the literature that the risk of respiratory infections may be raised. Because the AB can lead to increased intra-abdominal pressure a restrictive ventilation disorder due to diminished pulmonary volumes may result. Furthermore, it is assumed that the AB may lead to (transient) nerve palsies after prior transversus abdominis plane block, caused by atypical distribution of the local owing to the pressure exerted by the AB (25). It is also conceivable that amyotrophy is induced by a possible reduction in use of the abdominal muscles due to the AB (26). As a further possible negative effect of the AB, a survey showed that 32.6% of the patients who had undergone IH repair (n = 270) reported AB-induced immobility (11). In contrast to the potential negative impacts of the AB, we observed no postoperative complications among members of the AB group during their hospital stay in the pilot study reported here (CDC 0; eTable 2). In fact, albeit without significance, less limitation of mobility and higher general well-being was documented in the AB group than in the members of the control group.
The size of the postoperative seromas over time differed non-significantly between the two groups (p > 0.05; eTable 2, Figure 2). One person in the AB group suffered from a large seroma on postoperative days 1 (49 cm3) and 2 (49 cm3). Given the small sample size (n = 37), it is conceivable that this individual is responsible for the quantitative difference.
In this trial all individuals were operated on using the laparoscopic IPOM technique. This procedure has been described as adequate for the treatment of IH with closable gaps (circa < 7 cm) (27, 28). This may represent a limitation of the study, because most of the IH treated were less than 7 cm in size. The small sample size is a further limitation. The primary endpoint could be evaluated in only 12 of the 20 patients in the AB group, because eight individuals did not attend the second follow-up visit. Nevertheless, the findings of this pilot study allow power-calculation of sample size for further studies. As further limitations, we did not document the cumulative consumption of analgesics or their intake before each follow-up examination. Pain was measured only at rest using the VAS scale. Further studies should also record pain on exertion and cumulative analgesic consumption during the hospital stay and before all follow-up visits. The fact that the naturally examiner-dependent ultrasound investigation of the wound was conducted by different physicians must also be regarded as a limitation of the study.
Conclusion
The abdominal binder may lead to a reduction in pain following repair of incisional hernias using the IPOM technique.
Supplementary Material
eMethods
Methods
Descriptive statistics and mixed-effects model analysis
Descriptive statistics were used to characterize the patient group. Chi-squared tests were used to determine whether the two groups (abdominal binder vs. control) differed in categorical baseline characteristics and perioperative variables, including sex, ASA stage, hernia gap width, type of incisional hernia (IH) (primary/recurrent), gap closure, and surgeon’s experience. Independent t-tests were used to determine whether the two groups differed in continuous baseline characteristics and perioperative variables, including age, body mass index, operating time, and length of stay. For multiple comparisons, correction of statistical significance was achieved by means of the false discovery rate (FDR). Primary and secondary endpoints were compared between the two groups for the three postoperative days of interest (1, 2, and 14). Because repeated-measures analysis of variance (ANOVA) cannot handle missing values in the dependent variables, we analyzed the data with a mixed-effects model. To this end we used the software GraphPad Prism 8.0, in which the mixed-effects model utilizes a compound symmetry covariance matrix and is fitted using maximum likelihood methods. If there are missing values, the results can be interpreted like a repeated-measures ANOVA with group as between-subject factor and time (postoperative day 1, 2, and 14) as within-subject factor. Post-hoc analyses of time were adjusted using Holm–Šídák correction.
Sensitivity analysis
A sensitivity analysis was conducted with imputed data. SPSS (version 26) was used to replace missing data with multiple (n=5) imputations (regression model). The resulting F statistics were pooled using an approximation based on χ2 statistics using miceadds: : micombine.F in R (Version 4.0.2; R Core Team, 2020).
Footnotes
Conflict of interest statement
The authors declare that no conflict of interest exists.
Study registration
The study has been listed in the German Clinical Trials Registry (Deutsches Register Klinischer Studien, DRKS: ID DRKS00015665; www.drks-neu.uniklinik-freiburg.de/drks) and approved by the ethics committee of Berlin Medical Association (ETH-05/19). All patients gave written informed consent to study participation.
Funding
The clinical trial was funded by the Helios Research Center (HRC-ID 044537)
References
- 1.Hoer J, Lawong G, Klinge U, Schumpelick V. [Factors influencing the development of incisional hernia. A retrospective study of 2,983 laparotomy patients over a period of 10 years] Chirurg. 2002;73:474–480. doi: 10.1007/s00104-002-0425-5. [DOI] [PubMed] [Google Scholar]
- 2.Bittner R, Bingener-Casey J, Dietz U, et al. Guidelines for laparoscopic treatment of ventral and incisional abdominal wall hernias (International Endohernia Society [IEHS])-Part III. Surg Endosc. 2014;28:380–404. doi: 10.1007/s00464-013-3172-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bittner R, Bingener-Casey J, Dietz U, et al. Guidelines for laparoscopic treatment of ventral and incisional abdominal wall hernias (International Endohernia Society (IEHS)-part 1. Surg Endosc. 2014;28:2–29. doi: 10.1007/s00464-013-3170-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bittner R, Bingener-Casey J, Dietz U, et al. Guidelines for laparoscopic treatment of ventral and incisional abdominal wall hernias (International Endohernia Society [IEHS])-Part 2. Surg Endosc. 2014;28:353–379. doi: 10.1007/s00464-013-3171-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schwarz J, Reinpold W, Bittner R. Endoscopic mini/less open sublay technique (EMILOS)-a new technique for ventral hernia repair. Langenbecks Arch Surg. 2017;402:173–180. doi: 10.1007/s00423-016-1522-0. [DOI] [PubMed] [Google Scholar]
- 6.Bittner R, Bain K, Bansal VK, et al. Update of guidelines for laparoscopic treatment of ventral and incisional abdominal wall hernias (International Endohernia Society (IEHS)): Part B. Surg Endosc. 2019;33:3511–3549. doi: 10.1007/s00464-019-06908-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bittner R, Bain K, Bansal VK, et al. Update of guidelines for laparoscopic treatment of ventral and incisional abdominal wall hernias (International Endohernia Society (IEHS))-Part A. Surg Endosc. 2019;33:3069–3139. doi: 10.1007/s00464-019-06907-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chevrel JP. [The treatment of large midline incisional hernias by „overcoat“ plasty and prothesis (author‘s transl)] Nouv Presse Med. 1979;8:695–696. [PubMed] [Google Scholar]
- 9.Rives J, Pire JC, Flament JB, Palot JP, Body C. [Treatment of large eventrations. New therapeutic indications apropos of 322 cases] Chirurgie. 1985;111:215–225. [PubMed] [Google Scholar]
- 10.Antoniou SA, Agresta F, Garcia Alamino JM, et al. European Hernia Society guidelines on prevention and treatment of parastomal hernias. Hernia. 2018;22:183–198. doi: 10.1007/s10029-017-1697-5. [DOI] [PubMed] [Google Scholar]
- 11.Paasch C, Anders S, Strik MW. Postoperative-treatment following open incisional hernia repair: a survey and a review of literature. Int J Surg. 2018;53:320–325. doi: 10.1016/j.ijsu.2018.04.014. [DOI] [PubMed] [Google Scholar]
- 12.Paasch C, Lorenz E, Anders S, et al. Patient reported outcome following incisional hernia repair: a survey on 163 patients at two maximum care hospitals. Ann Med Surg (Lond) 2019;44:5–12. doi: 10.1016/j.amsu.2019.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bouvier A, Rat P, Drissi-Chbihi F, et al. Abdominal binders after laparotomy: review of the literature and French survey of policies. Hernia. 2014;18:501–506. doi: 10.1007/s10029-014-1264-2. [DOI] [PubMed] [Google Scholar]
- 14.Chowbey PK, Sharma A, Khullar R, Mann V, Baijal M, Vashistha A. Laparoscopic ventral hernia repair. J Laparoendosc Adv Surg Tech A. 2000;10:79–84. doi: 10.1089/lap.2000.10.79. [DOI] [PubMed] [Google Scholar]
- 15.Sharma A, Mehrotra M, Khullar R, Soni V, Baijal M, Chowbey PK. Laparoscopic ventral/incisional hernia repair: a single centre experience of 1,242 patients over a period of 13 years. Hernia. 2011;15:131–139. doi: 10.1007/s10029-010-0747-z. [DOI] [PubMed] [Google Scholar]
- 16.Team RC. R Foundation for Statistical Computing. Vienna, Austria: 2020,. A language and environment for statistical computing. [Google Scholar]
- 17.Bates DMM, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1–48. [Google Scholar]
- 18.Lüdecke D. sjstats: Statistical functions for regression models (Version 01.8.1) 105281/zenodo1284472 2021. [Google Scholar]
- 19.Molenberghs GVG. Models for discrete longitudinal data. Springer Series in Statistics Springer Journal. 2005 [Google Scholar]
- 20.Daniels MJ, Hogan JW. Missing data in longitudinal studies. Strategies for Bayesian modeling and sensitivity analysis, Chapman and Hall/CRC. 2008 [Google Scholar]
- 21.Wang F, Schilsky RL, Page D, et al. Development and validation of a natural language processing tool to generate the CONSORT reporting checklist for randomized clinical trials. JAMA network open. 2020;3 doi: 10.1001/jamanetworkopen.2020.14661. e2014661. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rothman JP, Gunnarsson U, Bisgaard T. Abdominal binders may reduce pain and improve physical function after major abdominal surgery—a systematic review. Dan Med J. 2014;61 A4941. [PubMed] [Google Scholar]
- 23.Christoffersen MW, Olsen BH, Rosenberg J, Bisgaard T. Randomized clinical trial on the postoperative use of an abdominal binder after laparoscopic umbilical and epigastric hernia repair. Hernia. 2015;19:147–153. doi: 10.1007/s10029-014-1289-6. [DOI] [PubMed] [Google Scholar]
- 24.Dart RC, Surratt HL, Cicero TJ, et al. Trends in opioid analgesic abuse and mortality in the United States. N Engl J Med. 2015;72:241–248. doi: 10.1056/NEJMsa1406143. [DOI] [PubMed] [Google Scholar]
- 25.Salaria ON, Kannan M, Kerner B, Goldman H. A rare complication of a TAP block performed after caesarean delivery. Case Rep Anesthesiol. 2017;2017 doi: 10.1155/2017/1072576. 1072576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vailas AC, Tipton CM, Matthes RD, Gart M. Physical activity and its influence on the repair process of medial collateral ligaments. Connect Tissue Res. 1981;9:25–31. doi: 10.3109/03008208109160237. [DOI] [PubMed] [Google Scholar]
- 27.Chelala E, Barake H, Estievenart J, Dessily M, Charara F, Alle JL. Long-term outcomes of 1326 laparoscopic incisional and ventral hernia repair with the routine suturing concept: a single institution experience. Hernia. 2016;20:101–110. doi: 10.1007/s10029-015-1397-y. [DOI] [PubMed] [Google Scholar]
- 28.Sajid MS, Bokhari SA, Mallick AS, Cheek E, Baig MK. Laparoscopic versus open repair of incisional/ventral hernia: a meta-analysis. Am J Surg. 2009;197:64–72. doi: 10.1016/j.amjsurg.2007.12.051. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods
Methods
Descriptive statistics and mixed-effects model analysis
Descriptive statistics were used to characterize the patient group. Chi-squared tests were used to determine whether the two groups (abdominal binder vs. control) differed in categorical baseline characteristics and perioperative variables, including sex, ASA stage, hernia gap width, type of incisional hernia (IH) (primary/recurrent), gap closure, and surgeon’s experience. Independent t-tests were used to determine whether the two groups differed in continuous baseline characteristics and perioperative variables, including age, body mass index, operating time, and length of stay. For multiple comparisons, correction of statistical significance was achieved by means of the false discovery rate (FDR). Primary and secondary endpoints were compared between the two groups for the three postoperative days of interest (1, 2, and 14). Because repeated-measures analysis of variance (ANOVA) cannot handle missing values in the dependent variables, we analyzed the data with a mixed-effects model. To this end we used the software GraphPad Prism 8.0, in which the mixed-effects model utilizes a compound symmetry covariance matrix and is fitted using maximum likelihood methods. If there are missing values, the results can be interpreted like a repeated-measures ANOVA with group as between-subject factor and time (postoperative day 1, 2, and 14) as within-subject factor. Post-hoc analyses of time were adjusted using Holm–Šídák correction.
Sensitivity analysis
A sensitivity analysis was conducted with imputed data. SPSS (version 26) was used to replace missing data with multiple (n=5) imputations (regression model). The resulting F statistics were pooled using an approximation based on χ2 statistics using miceadds: : micombine.F in R (Version 4.0.2; R Core Team, 2020).