Peritoneal antiseptic irrigation to prevent surgical site infection after laparotomy for hepatobiliary or gastrointestinal surgery (PAISI)—protocol for a randomized controlled study

Postoperative surgical site infection (SSI) is still one of the most common hospital infections. In Germany, approximately 200,000 SSIs are reported annually [1, 2]. In addition to the harm and discomfort for the patient, SSIs dramatically increase treatment costs and hospital occupancy. In Germany, postoperative SSIs account for approximately 1 million extra days of hospitalization and additional costs of around € 3 billion per year [3]. Other European countries report similar figures [4, 5]. According to the definition of the Centre for Disease Control and Prevention (CDC), SSI can be subdivided into 3 grades: class I (superficial wound infection), class II (deep wound infection) and class III (space/organ infection), see Table 1 [6]. While class I and II prolong the hospital stay and often require operative wound revision and long-term wound therapy, class III SSIs can be life-threatening and frequently require emergency relaparotomy, intensive care or interventional drainage of abscesses combined with systemic antibiotic therapy. Following open abdominal surgery with resections of the gastrointestinal or hepatobiliary tract, SSI rates remain especially high. Depending on the level of intraoperative contamination (LOC; see Table 2), recent high-level randomized controlled trials (RCTs) using standardized SSI definitions report postoperative SSI rates between 14.5 and 25.0% [8‐10]. However, in those trials the incidence of SSIs class III is rarely reported separately and is thus difficult to estimate. In our institution, we observed an overall SSI rate of 13%, of which 7% were SSI class III, following elective laparotomies for any reason (LOC II–III) from 2013 to 2018 (unpublished data).

Prophylactic intraoperative peritoneal irrigation (IPI) of the abdominal cavity before closure of the fascia with physiological electrolyte or antiseptic solutions hypothetically represents an easy and economical option to reduce SSI rates and is already frequently used in clinical practice. A survey among 153 general surgeons showed that 97% perform IPI, while the frequency of use and choice of lavage fluid varied widely [11]. However, the latest official guidelines for the prevention of SSI by the CDC (2017) [12] and the World Health Organization (WHO, 2016) [5] conclude, that the use of prophylactic abdominal or intra-incisional irrigation represents an unresolved issue, as the available evidence suggests uncertain benefits [5, 12]. In contrast, the recently updated British guideline recommends not to use any intracavity or wound irrigation to prevent SSI due to insufficient evidence for its benefits [7, 13]. A large-scale Cochrane meta-analysis of intra-cavity and wound irrigation for prevention of SSI which included 59 RCTs with 14,738 participants (most of which focused on orthopaedic and trauma surgery) identified only 3 RCTs which reported intra-abdominal abscess formation as a separate outcome. They concluded that current evidence shows that there is no clear difference in the rate of abscess formation with or without IPI [14]. However, the trials included in the analysis are very heterogenous, as they assessed between irrigation and no irrigation, between antibacterial and non-antibacterial irrigation, between different antibiotics, different antiseptics, or non-antibacterial agents, and between different methods of irrigation delivery or volumes. In a recently published consensus on the use of antiseptic irrigation solutions for prophylaxis of SSIs, it is stated that for peritoneal rinsing, sodium hypochlorite solutions (NaOCl/HOCl) are the agent of choice [15]. The bactericidal mechanism is mediated through the ion OCl-, which is also formed during phagocytosis through enzyme mediation by myeloperoxidase, eosinophilic peroxidase, and superoxide dismutase [15, 16]. Pre-clinical studies, which were reviewed for the consensus showed that NaOCl/HOCl is highly effective against vegetative bacteria, bacterial spores, aspergilli, oocysts of cryptosporidia, coated viruses and against biofilm [15]. In addition, the speed of effect is superior to PVP-iodine, Octenidine and Polyhexanide. Due to its physiological mechanism of action, there is no evidence for any undesirable effects, cytotoxicity, or carcinogenic hazard [15].

SSIs contribute significantly to postoperative morbidity and mortality. Clinical guidelines and clinical practice vary largely in terms of the use of peritoneal and wound irrigation to reduce the incidence of SSI [24]. The PAISI study aims to investigate if IPI with a NaOCl/HOCl solution can reduce the rate of postoperative SSI within 30 days (according to the CDC definition shown in Table 1) after resections of the gastrointestinal or hepatobiliary tract by laparotomy. The results of this study will provide high-level evidence for future research and clinical recommendations regarding the use of NaOCl/HOCl solution in abdominal surgery and provide the participating patients the opportunity of a potentially improved treatment.

This is a prospective, randomized, controlled, observer and patient-blinded, single-centre, two-arm surgical study with an adaptive parallel groups design. Pre-screening of potential patients (evaluation of inclusion and exclusion criteria) is possible up to 14 days prior to the planned procedure. Patients can be included if inclusion and exclusion criteria apply and written informed consent is provided. Included patients are randomized to peritoneal and epifascial wound irrigation with NaOCl/HOCl (50/50ppm) or Ringer solution. A reduction of 30-day postoperative SSI I–III rates by lavage with NaOCl/HOCl solution is postulated. Since there is currently no data on incidence rates from randomized clinical trials, the rates for sample size calculation were estimated according to the clinical experience at our institution. Therefore, the study is designed with one unblinded look at the data, which will be performed after half of the patients reached the primary endpoint (30 days post-surgery), by a second study statistician, who will not participate in the final study analysis. This interim information will be used to check the assumptions which were used to calculate the sample size for the study. If needed, the sample size will be adjusted.

The PAISI study will be conducted in the surgical department of the “Klinikum Rechts der Isar” of the Technical University of Munich, which is a member of the German trial network (CHIR-Net) of the German Surgical Society (Deutsche Gesellschaft für Chirurgie) and has extensive experience regarding studies on prevention of SSIs. All study personnel involved has adequate training and will be instructed in all specific procedures before the initiation of the study. The leading surgeon of the operating team will perform the intervention since it represents a routine technique. All participating surgeons will be instructed and authorized by the investigator, prior to the first study procedure.

A patient can only be included in the study, if he provides written consent after being informed by a clinical investigator (orally and in writing) about the nature, significance, and scope of the clinical study in an appropriate and understandable way. Clinical investigators are specifically trained medical doctors of the local study team. The investigator must fully explain the purpose of the study to the patient or his/her guardian prior to entering the patient into the study. The investigator is responsible for obtaining written informed consent from each patient.

Not applicable as no biological specimens are collected, and the participant data will not be used for other purposes than the study. The IC form contains a data protection statement regarding the consent of patient must be given that relevant data can be used for the study in a pseudonymous form.

The IPI with physiological electrolyte solution (e.g. Ringer’s solution) at the end of abdominal surgical procedures represents a current clinical routine procedure. In many hospitals, even IPI with NaOCl/HOCl (50/50ppm) solutions is already considered the gold standard for septic or contaminated procedures. However, most hospitals do not have standard protocols but leave the decision if, how and when to irrigate the abdominal cavity and/or the incisional wound up to the surgeon. However, in current national and international guidelines, there are no specific recommendations on the performance, volume or timing of IPI [5]. Modern antiseptic solutions like NaOCl/HOCl or Polyhexanide are not mentioned in the current guidelines.

No additional risks for study patients are anticipated since IPI represents a clinically established standard method. NaOCl/HOCl (50/50ppm) and physiological Ringer’s solution are CE-certified for peritoneal and surgical wound irrigation of soft tissue wounds. The study will be planned, conducted, and analysed according to all relevant rules and regulations and the Declaration of Helsinki, 2008. No specific risks are expected because neither application of NaOCl/HOCl nor Ringer’s solution will have systemic effects on the participants. Adverse effects may only be expected in the improbable event of accidental contamination of the respective irrigation solutions. The potential benefits of reduced SSIs outweigh the mentioned negligible potential adverse effects. The informed consent process adheres to Good Scientific Practice, which maximizes patients’ safety and confidentiality.

At the end of surgery, before the closure of the abdominal fascia, patients will be randomized stratified by LOC of the operation. In the experimental group, the abdominal cavity will be irrigated with 1000 ml NaOCl/HOCl (50/50ppm) solution. In addition, after closure of the fascia, the wound shall be carefully rinsed throughout with another 500 ml of the irrigation solution and the excess removed with suction. Debris and blood clots should be removed from the wound using irrigation/suction. The wound shall not be rubbed dry with abdominal cloths but left moistened with the irrigation solution to ensure the desired antiseptic effect. After irrigation with NaOCl/HOCl, the abdominal cavity and wound shall not be irrigated with any other solution again.

In the control group, the same intervention will be performed using 1000 ml of Ringer’s solution for the abdominal lavage and 500 ml of the same solution for the incisional wound (current standard). The irrigation volume of 1500 ml was chosen to be sure that the abdominal cavity and even large laparotomy wounds would be sufficiently irrigated. This was determined by senior surgeons´ clinical experience, since so far, no recommendations for the optimal volume of surgical irrigation exist. Closure of the fascia and the skin will be performed according to local standards, without any further wound-related procedure.

NaOCl/HOCl (50/50ppm) and Ringer’s solutions are purchased, stored, and distributed according to the local standard operating procedures. Trade name, dosage, batch and dispensed amount will be documented on a separate form.

Participants of the study can withdraw their consent to take part at any time without declaration of reasons. All hitherto collected data are subject to analysis, within the consent of the participant. The principal investigator may exclude patients from the study, if patients’ safety is at risk or if there is insufficient compliance of the patient. In order to generate a meaningful database, excluded patients can be replaced by recruitment of new patients. If a patient does not receive NaOCl/HOCl or Ringer solution irrigation of the peritoneal cavity or wound, this does not automatically lead to exclusion of the study.

Not applicable as the intervention is performed only once, intraoperatively. Adherence to the follow-up visit schedule is promoted by facilitating the last two study visits 10 and 30 days postoperatively as a structured telephone interview. Furthermore, all study procedures and visits closely orientate towards standard medical care.

No additional treatments or interventions will be performed within the study. All patients will receive standard perioperative antibiotic prophylaxis, the timing and type of which will be recorded in the eCRF. Furthermore, the application of abdominal wall protectors is recommended for contaminated procedures and must be recorded in the eCRF. A change of gloves ahead of wound closure is frequently performed in contaminated procedures and must also be recorded in the eCRF. If an irrigation-suction-drain is placed in the abdominal cavity and the patient receives prophylactic intraabdominal irrigation therapy via this device postoperatively, the patient must be excluded from the study. If indicated for medical reasons, all kind of medication is permitted during the study. Postoperative medication with known adverse effects on wound healing (e.g. antibiotics, corticoids, and other immunosuppressive agents) will be recorded in the eCRF. Any operative or interventional revision of the wound will be documented as AE/ SAE in the eCRF.

Not applicable as no additional risks and harms from study participation are anticipated, and no additional travel is necessary for the patient.

The primary efficacy endpoint of the study is occurrence of SSI within 30 days postoperatively, according to the internationally accepted and recommended SSI definition by the CDC (Table 1) [6]. This endpoint has been used in previous studies and assures comparability of the results [8‐10, 25]. This endpoint is further considered to be of clinical relevance as SSI increases morbidity and mortality of individual patients, direct and indirect costs and prolongs hospital stay as outlined before. The duration of hospital stay is considered as a secondary outcome, in order to evaluate the potential economic benefit of the intervention. In addition, the following outcome measures have been defined as secondary endpoint measures and will be determined by the unit given in parentheses:

Figure 1 and Table 3 reflect the intervention scheme flow and visits for the PAISI study. Visits are the same for all participants of the study, regardless of the intervention group.

The sample size was calculated (nQuery 8 software, version 8.6.1.0, Statistical Solutions Ltd, Cork, Ireland) based on the primary endpoint of the study. Assuming SSI rates of 4% and 16% for the experimental and the standard group respectively (OR=4.571), the one-sided z-test (Group Sequential test of two proportions) at the 2.5% significance level with a total of two looks at the data will have overall 80% power to show superiority of the experimental over the standard arm if the study includes at least 232 patients (116 patients per intervention group). The calculation is continuity corrected due to the expected low proportion in the experimental group. The results assume that the group sequential design has one interim sequential test for monitoring and sample size re-estimation (two total looks at the data including final analysis), which means that an interim analysis is planned after half of the patients have reached 30 days post-surgery. The O’Brien-Fleming spending function is used to determine the efficacy test boundary and the non-binding futility boundary. The group sequential test boundaries (Z scale) for the interim and the final analysis are as follows:

Upper efficacy bounds: 2.963 and 1.969

Futility bounds: 0.555 (non-binding) and 1.1969.

The recruitment period is expected to be approximately 32 months (first patient in, to last patient out 33 months). In the study site, an average of around 80 laparotomies LOC II–IV are performed per month. Experience from previous clinical studies shows that the recruitment of 10 patients per month is feasible here. To cover any unforeseen recruitment difficulties, for example due to the COVID-19 pandemic, 3 extra months of recruitment time were added. In the case of elective procedures, pre-screening (this is just a pre-selection of eligible patients within the study team) of patients can be performed up to 14 days prior to the scheduled surgical procedure. In case of emergency procedures, screening and inclusion can take place on the day of admission to the hospital, which is usually the same day as surgery.

The randomization list will be generated by an online randomization tool (Randomizer, Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Austria; URL: http://​www.​randomizer.​at). Online-Randomization will be performed at the end of the surgical procedure before the abdominal fascia is closed and will be stratified by LOC of the surgical procedure (LOC II–IV; Table 2). To assure balanced group sizes during the accrual, a block-wise randomization is applied.

Only designated members of the study team, who will not perform postoperative study visits, will have access to the online randomization tool. After informing the surgical team of the randomization result, the randomization sheets will be printed out and must be stored away from the patient records, study documents and ISF to ensure blinding of the rest of the local study team.

The allocation sequence is generated by the Randomizer tool (Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Austria; URL: http://​www.​randomizer.​at). Participants will be enrolled by the specifically trained medical doctors of the local study team. Participants will be assigned to the intervention groups by the designated study nurse or medical doctor who is effectuating the online randomization.

The blinding procedure is restricted to participating patients, the statistician who performs the final analysis and the outcome assessors. Blinding of the surgical team that performs the intervention is impossible because of the distinctive smell of NaOCl/HOCl solution. Postoperatively, an outcome assessor of the local study group, who is unaware of the patient’s intraoperative intervention, will clinically assess the primary endpoint (SSI) on 3 postoperative study visits.

Unblinding of outcome assessors is permissible if significant hazards for patients’ safety or welfare occur. In such an unlikely case, the member of the study team responsible for the randomization procedure can unblind the outcome assessor for the randomization result of the concerned patient, only.

To promote complete follow-up, a visit window of two additional days was implemented for visit 3, and 5 additional days for visits 4 and 5. If however, the patient is unable to attend visit 4 or 5 due to postoperative treatment in a rehabilitation facility or other medical reasons, a standardized protocol for evaluation and documentation of the wound or intraabdominal complications will be sent to and filled out by the treating physician. In the exceptional case that the patient has no possibility to be evaluated by a physician for visit 5, a telephone interview with the patient can be performed to assess the primary endpoint. However, in this case, the reasons why a physician cannot see the patient have to be declared on a designated form.

The documentation of the study data in adherence to the protocol is the responsibility of the principal investigator. Original data (source documents) remain in hospital medical record and information on the eCRF must be traceable and consistent with the original data. Original written informed consent signed by the patient is kept by the investigator and a signed copy will be given to the patient. No information in source documents about the identity of the patients will be disclosed. All data collected in this study must be entered in an eCRF using a REDCap database (Research Electronic Data Capture, Version 10.5.2, 2022, Vanderbilt University) [30, 31] which has to be completed by the investigator or authorized study personnel and signed by the investigator. This also applies for those patients who do not complete the study. If a patient is excluded from the study, the reason must be recorded on the eCRF. The investigator is responsible for ensuring the accuracy, completeness, and timeliness of all data in the eCRFs and in all required reports. After entry of all collected data and clarification of all queries, the database will be closed at the completion of the study. Data and results electronically recorded will be archived according to applicable law.

The applicable local regulations of data privacy protection will be followed. The patients will be informed that any patient-related data and materials will be appropriately made pseudonymous and that these data may be used for analysis and publication purposes. Furthermore, the patients will be informed that their data may be inspected by monitors for the purpose of validation of a proper study conduct. Patients who do not provide consent for processing of their data, according to the data protection agreement included in the ICF, will not be included in the clinical study.

Not applicable since no biological specimen will be taken.

The intention-to-treat (ITT) set consists of all patients who received irrigation of the abdominal cavity before closure of the abdominal fascia with one of the study solutions. Analysis will be performed as randomized regardless of the actual intervention and regardless of any further medical or surgical treatments. The Per-Protocol (PP) set consists of patients treated according to protocol, excluding any major protocol violations. The PP set will be analysed as treated. The Safety (SA) set will consist of all patients in the ITT set, analysed as treated.

All primary and secondary analyses will be performed on the ITT set. Sensitivity analysis of the primary endpoint will be performed on the PP set. The SA set will be used for the safety analyses. The incidence of SSI within 30 days after surgery will be compared between the two study groups using the one-sided group sequential test of two proportions at the 2.5% significance level with a total of two looks at the data. The primary endpoint analysis will be performed on the ITT set. Since the ITT set includes all patients who were treated in the study regardless of any further medical or surgical treatments, it is expected, that the SSI rates 30 days post-surgery will be available for most patients. Nevertheless, it is possible that some patients do not reach the primary endpoint due to death, withdrawal of informed consent, loss to follow-up, or other unforeseen reasons. Patients who die before day 30 will be considered as having SSI at time of death regardless of the intervention group, unless SSI was documented earlier. All other missing primary endpoint data will be dealt with using multiple imputations based on binary logistic regression models including LOC, ASA, BMI, age, diabetes, type of surgery, duration of surgery, use of wound-edge protectors, and intraoperative changing of gloves.

Thirty days after the first half of the patients has reached the primary endpoint, an interim analysis will be performed by a second statistician, who does not take part in the final study analysis. A decision will be made by the study leadership, using the blinded results to stop the study with proven efficacy, stop the study for futility, continue the study as planned, or continue the study after adjusting the sample size.

Since randomization will be stratified by LOC, supportive analysis of the primary endpoint will be performed using a binary logistic regression model with dependent variable SSI and covariates intervention group and LOC. In addition, the following parameters might influence the outcome, which is why they will be included as model covariates in an additional model:

Supportive analysis of the primary endpoint will be performed on the ITT set using multiple imputations for any missing SSI values and on the PP set. Secondary endpoints will be analysed by intervention arm on the ITT set, using appropriate descriptive statistics. Any explorative statistical testing will be performed using a significance level of 5%. The assessment of safety will be based on the frequency of AE/SAE as defined in the protocol within the safety population consisting of all patients randomized into the study.

Multiple imputations will be used for missing primary endpoint data. All other data will be analysed on a complete-case basis.

The protocol publication will be accessible on the DRKS website (www.​drks.​de). This includes a summary in lay language for the interested patient community. The full protocol and statistical code can be requested from the authors. Individual participant level data that underline the results reported in this article can be shared after anonymization, with investigators whose proposed use for the data has been approved by an independent review committee identified for this purpose. Proposals can be submitted up to 36 months after publication of the study results.

As this study is planned as a monocentric trial, there is no coordinating centre. The trial steering committee in this setting consist of the study leadership (principal investigator, deputy investigator and coordinating investigator) which is responsible for recruitment of patients and data management. The statistical analysis will be performed by an independent statistician. Trained medical doctors of the local study group will take informed consent. The study leadership will meet once a month to discuss progress and problems regarding the trial. Following the planned interim analysis after 50% of patients are recruited a meeting with the trial statistician and the scientific advisors as well as the funders will be held to discuss the results and consequences. The ethics committee will be informed about the outcome of this meeting. Public involvement is not planned in this study.

In this study setting, a data monitoring committee is not planned, as the intervention poses no additional risks for the patients and is a routine medical procedure.

Each adverse event (AE) is to be classified by the investigator according to the Common Terminology Criteria CTCAE V 5.0 [32]. Serious AEs and AEs of special interest are captured and processed until the last visit of the last patient. A final list will be provided to the ethics committee. All documented (S)AEs should be followed to resolution or stabilization. The investigator will be requested to supply as much detailed information as possible regarding the event that is available at the time of the initial contact. The investigator is also required to complete missing or requested information until the (S)AE has resolved or, in the case of permanent impairment, until stabilized. All (S)AEs severity will be additionally graded by the Clavien Dindo Classification of postoperative complications and the CCI.

As this study is comparing two routine surgical procedures, auditing of the trial by external institutions is not required and therefore not planned.

Any amendments to the protocol, other than administrative ones (of which the ethics committee will merely be informed), must be reviewed and approved by the ethics committee. The changes will be updated in the clinical register DRKS on a regular basis. If necessary, patient information is updated to inform new participants of the changes.

After completion of the clinical study, a manuscript of the study results will be prepared for publication in a reputable scientific journal according to the CONSORT statement. For this manuscript, final analyses will be generated from the study database, and it will be subject to review by the principal investigator. The use of professional writers is not intended.

The first patient was recruited in September 2022, therefore recruitment is planned to be completed approximately by May 2025.