Postoperative Outcomes Within an Enhanced Recovery After Surgery Protocol

Overview

Short title POWER Audit Methods 60 days national (Spain) audit of postoperative complications following elective gastrointestinal surgery within an enhanced recovery after surgery (ERAS) protocol. Research sites Hospitals undertaking elective lower gastrointestinal surgery. Objective To provide detailed data describing post-operative complications and associated mortality; and length of stay. To provide detailed data describing adherence to ERAS protocol and its association to morbidity. Number of patients Not specified. All eligible patients undergoing surgery during the study month. Inclusion Criteria All adult patients (aged ≥18 years) undergoing lower gastrointestinal elective surgery within an ERAS protocol during the 60 day study period. Statistical analysis Univariate analysis will be used to test factors (patient, surgical, and ERAS related) associated with surgical complications, length of stay (LOS) and in-hospital death. Single and multi-level logistic regression models will be constructed to identify factors independently associated with these outcomes and to adjust for differences in confounding factors. A stepwise approach will be used to enter new terms. A single final analysis is planned at the end of the study. Summary statistics with post hoc Bonferroni corrections will be used to assess possible dose-response dependence in percentage of patients with postoperative complications and LOS. Proposed Start Date A 60 day period between 2017 Proposed End Date Data collection will end by September 2017 Study Duration Six months

Study data Data will be collected on all eligible patients who undergo elective lower gastrointestinal surgery within any compliance of an ERAS protocol during the study months. Only routine clinical data will be included and where this is unavailable the domain will be left blank e.g. patients who do not require blood tests. It is possible that regional groups may supplement their core data set with a very limited number of additional variables if these can be accommodated within the two page case record form (CRF) and they comply with regulations applied to this study. Data collection Data will be collected in individual hospitals on a paper CRF for each patient recruited. Paper CRFs will be stored within a locked office in each centre. This will include identifiable patient data in order to allow follow-up of clinical outcomes. Data will then be pseudo-anonymised by generating a unique numeric code and transcribed by local investigators onto an internet based electronic CRF. Each patient will only be identified on the electronic CRF (eCRF) by their numeric code. Thus the co- ordinating study team cannot trace data back to an individual patient without contact with the local team. A patient list will be used in each centre to match identifier codes in the database to individual patients in order to record clinical outcomes and supply any missing data points. Once the local co-ordinator confirms data entry is complete for their hospital they will receive a spreadsheet of raw (un-cleaned) data, allowing further checks for data completeness and accuracy. Audit organisation POWER will be led by the study management group who will be responsible for study administration, communication between project partners, data collation and data management. Regional co-ordinators will lead the project in each Spanish region and: * Identify local co-ordinators in participating hospitals * Assist with translation of study paperwork as required * Ensure distribution of study paperwork and other materials * Ensure necessary regulatory approvals are in place prior to the start date * Ensure good communication with the participating sites in his/her region Local co-ordinators in individual institutions will have the following responsibilities: * Provide leadership for the study in their institution * Ensure all relevant regulatory approvals are in place for their institution * Ensure adequate training of all relevant staff prior to data collection * Supervise daily data collection and assist with problem solving * Act as guarantor for the integrity and quality of data collected * Ensure timely completion of eCRFs by supervising local data entry * Communicate with the relevant national co-ordinator End of Study Definition The end of the study is defined as the end of the 30-day follow-up for the last patient included. Data analysis shall follow this. STATISTICAL ANALYSIS Sample size calculation Our plan is to recruit as many centres as possible on an international basis and ask them to include all eligible patients in the study. Only centres including 10 valid patients will be included in the data analysis. We do not have a specific sample size and statistical models will be adapted to the event rate provided by the sample recruited. However, a minimum sample size is estimated, expecting 50% of patients with at least one complication, which is the data that requires a higher sample-size with a confidence level of 95% and an accuracy of 3%, a total of 1068 patients. A larger sample size, greater accuracy will be achieved. Statistical analysis Hospitals including data describing less than 10 valid patients will be excluded in the data analysis. Data will be presented in individual regions. All regional and institutional level data will be anonymised prior to publication. Categorical variables will be described as proportions and will be compared using chi-square or Fisher's exact test. Continuous variables will be described as mean and standard deviation, if normally distributed, or median and inter-quartile range, if not normally distributed. Comparisons of continuous variables will be performed using one-way ANOVA or Mann-Whitney test as appropriate. Univariate analysis will be performed to test factors associated with post-operative complications, LOS and in-hospital death. Single-level and hierarchical multi-level logistic regression models will be constructed to identify factors independently associated with these outcomes and to adjust for differences in confounding factors. Factors will be entered into the models based on their univariate relation to outcome (p\<0.05), biological plausibility and low rate of missing data. A stepwise approach will be used to enter new terms. Results of logistic regression will be reported as adjusted odds ratios (OR) with 95% confidence intervals. The models will be assessed through the use of sensitivity analyses to explore possible interacting factors and examine any effect on the results. A single final analysis is planned at the end of the study. Overall compliance will be calculated as the average of all pre- and intraoperative ERAS adapted elements, as specified in the ERAS society colon and rectal guidelines. For exploratory purposes, post-ERAS patients' guideline compliance will be categorised into quintiles: compliance \<45% (compliance 1), compliance 45-55 % (compliance 2), compliance 55-65 % (compliance 3), compliance 65-75 % (compliance 4) and compliance \>75 % (compliance 5). Summary statistics with post hoc Bonferroni corrections will be used to assess possible dose-response dependence in percenctage of patients with postoperative complications and LOS. The data set will be analyzed using the percentage of patients with postoperative complications and LOS the main and secondary outcome variables. The influence of the following factors was assessed: sex, age, American Society of anesthesia (ASA) status, body mass index (BMI), preoperative hemoglobin, comorbidity, including hypertension , diabetes mellitus, coronary arterial disease , chronic obstructive pulmonary disease, and chronic renal disease; surgical approach (open, laparoscopic), duration of surgery, intraoperative fluid administration and first 24 hours fluid balance; and individual components of the ERAS protocol. Univariate analysis will be initially undertaken to assess the relationship between each factor and the outcome variables. Comparisons will be made using the χ2 test for all categorical variables and the t test and Kruskal-Wallis test will be used to evaluate differences between continuous normally and non-normally distributed variables, respectively. Owing to its non-normal distribution, LOS will be analyzed by log-normal transformation and independent t tests with back exponentiation. Multivariate analysis, using binary logistic regression for development of complications and linear regression of log transformed length of LOS, will be then performed for all variables in the univariate analysis with a significant or near-significant difference (P \< 0.1). P \< 0.05 will be considered statistically significant. Primary outcome measure The percentage of patients who developed pre-defined mild-moderate-severe postoperative complications in the 30 days after surgery, including complications that occurred before hospital discharge and those that happened after discharge and required ambulatory or in-hospital care. Postoperative complications are defined according standards for definitions and use of outcome measures for clinical effectiveness research in perioperative medicine (EPCO). Secondary outcome measures * In-hospital all-cause mortality (censored at 30 days following surgery) * Adherence to ERAS items (within 30 days following surgery) * Duration of hospital stay (duration of primary hospital stay after surgery) ETHICS The principal investigator must ensure that the study will be carried out in accordance with the ethical principles in the Research Governance Framework for Health and Social Care, Second Edition, 2005 and its subsequent amendments as applicable and applicable legal and regulatory requirements. Research ethics approval may not be required in all participating nations. Regional and local investigators will be responsible for clarifying the need for ethics and other regulatory approvals and for ensuring these are in place prior to data collection. Centres will not be permitted to record data without providing confirmation that the necessary ethics or other regulatory approvals are in place. There will be no requirement for individual patient consent as all data will be anonymised and are already recorded as part of routine clinical care. SAFETY CONSIDERATIONS There are no safety considerations relating to the POWER Audit . There is no risk of harm to either patients or investigators. DATA HANDLING AND RECORD KEEPING All identifiable data collected, processed and stored for the purposes of the project will remain confidential at all times and comply with Good Clinical Practice for research (GCP) guidelines and the principles of the Data Protection Act 1998 (UK). Each centre will maintain a trial file including a protocol, local investigator delegation log, documentation of the relevant regulatory approvals and patient list. POWER data collection sheets will be stored securely in a locked cupboard and handled only by clinical staff familiar will handling personal data and with Good Clinical Practice for research. Data will be anonymised prior to transfer to the POWER Audit management group except where the patient has given written informed consent to allow transfer of identifiable data. Access to the data entry system will be protected by username and password, delivered during the registration process for individual local investigators. All electronic data transfer between participating centres and the coordinating centre will be encrypted. Desktop and laptop security will be maintained through user names and passwords. All local investigators will be asked to undergo training in accordance with the Research Governance Framework. The study master files will be stored in an approved repository for 20 years following the end of the study SAFETY REPORTING The trial involves negligible risks to patients and investigators. Adverse events will not be monitored or reported. MONITORING \& AUDITING POWER Audit master documents will be audited by the sponsor ("Grupo Español de Rehabilitación Multimodal", (GERM)) to ensure study activities are conducted according to the protocol, the sponsor's standard operating procedures, Good Clinical Practice and the applicable regulatory requirements. In participating hospitals, local study documents may be selected for audit on a local basis. However, the POWER Audit team will not routinely monitor data collection in individual hospitals or conduct source data verification. TRIAL COMMITTEES Trial Management Group The POWER Audit will be managed by the GERM Network and by the "Evidence Anesthesia Review Group (EAR)" team based at Complutense University of Madrid. The day-to-day conduct of the trial will be led by the trial management group, chaired by Javier Ripollés-Melchor. Trial Steering Committee The trial steering committee will be appointed with an independent chairperson (José Manuel Ramírez), lay representation and independent members. There is no role for a Data Monitoring Committee. FINANCE AND FUNDING The POWER Audit is funded by an unrestricted research grant from GERM Network ("Grupo Español de Rehabilitación Multimodal"). The funder will play no role in study design, conduct, data collection, data analysis, reporting or interpretation of the results. INDEMNITY The POWER Audit is sponsored by GERM Network ("Grupo Español de Rehabilitación Multimodal") who has appropriate indemnity arrangements in place. DISSEMINATION OF RESEARCH FINDINGS The steering committee will appoint a writing committee to draft the scientific report(s) of this investigation, which will be disseminated in a timely manner. It is anticipated that a number of secondary analyses will be performed. POWER investigators will be given priority to lead such analyses and are encouraged to do so. Participation and authorship opportunities will be based on contribution to the primary study. The steering committee will consider the scientific validity and the possible effect on the anonymity of participating centres prior to granting any such requests. Where necessary, a prior written agreement will set out the terms of such collaborations. The steering committee must approve the final version of all manuscripts including POWER data prior to submission. In the event of disagreement within the steering committee, the chief investigator will make a ruling. Any analysis incorporating POWER data from two or more study sites will be considered a secondary analysis and subject to these rules. The e CRF will provide local co-ordinators with the raw (un-cleaned) data for their centre once they have confirmed this to be both complete and accurate. Data management and ownership The study sponsor, "Grupo Español de Rehabilitación Multimodal", will act as custodian of the data. In line with the principles of data preservation and sharing, the steering committee will, after publication of the overall dataset, consider all reasonable requests to conduct secondary analyses. The primary consideration for such decisions will be the quality and validity of any proposed analysis. Only summary data will be presented publicly and all national, institutional and patient level data will be strictly anonymised. Individual patient data provided by participating hospitals remain the property of the respective institution. Once each local co-ordinator has confirmed the data provided from their hospital are both complete and accurate, they will be provided with a spreadsheet of the raw (un-cleaned) data for their hospital. The complete POWER dataset, anonymised with respect to participating patients, hospitals and nations, will be made freely and publicly available two years following publication of the main scientific report. Prior to this, the steering committee is not under any obligation to release data to any collaborator or third party if they believe this is not in keeping with the wider aims of the POWER Audit.