The main objective of this Italian nationwide study is to characterize the clinico-pathological findings, management strategies and short-term outcome of elderly patients undergoing emergency surgery. The secondary objectives are to evaluate the presence and influence of frailty, and to analyze the prognostic role of existing risk scores to define the most suitable scoring system to classify the elderly. We will conduct an epidemiological investigation to gather information about the number of elderly patients operated on yearly, and the prevalence of various pathological conditions leading elderly patients to need emergency surgery. Furthermore, we aim to develop and validate a new simplified modified Frailty Index suitable for emergency surgery. The study was registered at ClinicalTrials.gov (ClinicalTrials.gov Identifier: NCT02825082).

The ERASO (Elderly Risk Assessment And Surgical Outcome) Collaborative Study Group, on behalf of the Italian surgical societies ‘Italian Society for Emergency and Trauma Surgery’ (Società Italiana di Chirurgia d'Urgenza e del Trauma, SICUT), ‘Italian Hospital Surgeons Association’ (Associazione dei Chirurghi Ospedalieri Italiani, ACOI), ‘Italian Society for Geriatric Surgery’ (Società Italiana di Chirurgia Geriatrica, SICG), and the Italian Chapter of the ‘World Society of Emergency Surgery’ (WSES), designed the FRAILESEL study. The prospective, observational study will be conducted following a resident-led model, similar to what has been described by Banghu et al. ²⁶ and vanRossem et al ²⁷ . Briefly, teams of medical students and surgical residents with senior staff surgeon oversight collect data on patients across Italy for 18 months. The Data Coordination Center (DCC) is the Emergency Surgery Unit of Sant'Andrea Hospital, one of the teaching hospitals of "Sapienza" University of Rome, and the Study Director is responsible for the selection of the study sites. Any center performing emergency surgery can participate in this trial. Twenty-eight centers currently participate in the study and have started enrolment. The centers include academic medical centers, teaching hospitals, tertiary referral centers and community hospitals. Investigators involved in the study were trained in clinical score calculation prior to initiation of the study. Furthermore, all centers and participants have been specifically instructed in searching for frailty criteria and in data collection methodologies. Frailty is assessed by considering the frailty phenotypes as descried by Fried et al. ²⁸ To ensure that there is a uniform protocol of data acquisition in place for the main study, a 3-month pilot study was performed, that collected data retrospectively. Such data are not yet published, and are stored separately for further analyses. No modifications to the study protocol or database were necessary after the pilot phase. All patients are treated according to the local hospital protocol, and receive routine care as standard therapy. The duration of the recruitment phase of patients is expected to be 18 months. The main strength of this project is the multicenter, prospective, contemporary methodology, with independent validation of data. This will produce high quality data on the emergency procedures carried out in the elderly, and on outcomes throughout Italy from a wide range of hospital types. Limitations include the inability to assess the postoperative visits to the general practitioner. Moreover, a minority of patients may present to other hospitals with complications following surgery, or because they need medical assessment. Despite this, teams will try to document the number of patients that were readmitted to other facilities. Our study uses the standard 30 day follow-up period, as this is the international standard and allows comparison to other studies. However, complications which may occur after 30 days will be reported.

Inclusion criteria: All patients at the age of 65 years or over who are undergoing emergency surgery can be entered into this study. Emergency procedures are defined as unforeseen, non-elective operations according to the NCEPOD Classification of Interventions ²⁹ . The type of surgical approach takes into account open abdominal or laparoscopic procedures, including laparoscopic procedures that are converted to open abdominal procedures. Surgical procedures will be sorted on the basis of the 9 ^th revision of International Classification of Disease Clinical Modification (ICD-9-CM). All abdominal procedures with ICD-9-CM code numbers ranging from 42.0 to 54.99 are considered eligible. Thoracic procedures [ICD-9-CM code 32.0 to 34.99], vascular procedures [ICD-9-CM code 38.0 to 39.99], gynecological procedures [ICD-9-CM code 55.0 to 59.99], and urological procedures [ICD-9-CM code 60.0 to 64.99] are considered eligible for the study if performed by general or emergency surgeon in a general or trauma surgery setting.

Exclusion criteria: Exclusion criteria include patients aged under 65 at the day of surgery; lack of informed consent; patients already hospitalized and scheduled for the same procedure; participation in another trial.

The primary outcomes are 30-day overall postoperative morbidity and mortality rates. Secondary outcomes are 30-day postoperative morbidity and mortality rates, stratified for each procedure or cause of intervention, length of hospital stay, admission and length of stay in ICU, and place of discharge (home or rehabilitation or care facility). Other secondary outcomes include the number of elderly subjects undergoing yearly emergency surgery, reported as the elderly to non elderly patient ratio, emergency surgery in the elderly per 100.000 inhabitants, sensitivity, specificity, and the overall predictive value for morbidity and mortality of P-POSSUM, CR-POSSUM, SAPS II, mFI and CACI. The postoperative complications are reported and categorized according to the Clavien-Dindo classification system.

In each participating hospital, one local investigator (usually a surgical resident) is responsible for data collection and for entering data into a password-protected electronic spreadsheet (Excel 2010; Microsoft, Redmond, Washington, USA) specifically constructed with predefined data fields. There are six categories, namely “patient demographics”, “comorbidities”, “clinicopathological data”, “surgical intervention”, “score”, and “follow-ups” ( Table 1). Patient details will be recorded and anonymized using the code center, an ID number and a unique alphanumeric code for any further integration. The anonymization procedure is provided by the enrolling center. Patient data will be collected, if possible, on a daily basis; preoperative and intraoperative data will be processed after surgery, and the postoperative outcomes will be noted at the time of discharge and at the end of follow-ups. Data will be obtained from the electronic patient database, from admission charts, and operative reports, or directly from the surgeon who performed the operation when details were unclear or missing. Consent to participate in the study and to collect data for scientific purpose will be obtained from the patient at admission. The standardized data collection protocol has been approved by the Central Ethical Committee. There is no minimum number of patients per center. Following data collection, only datasets with >95% data completeness will be accepted for pooled national analysis. The principal investigator (PI) at the selected site will identify an independent assessor to validate all data, with a target of >98% accuracy. Overall, at least 5% of the datasets will be independently validated. Outcome data will not be analyzed specific to each individual center. Data will be submitted monthly via e-mail. Once in the Data Coordination Center (DCC) pooled warehouse, records are reviewed and edited and, whenever necessary, transformed to comply with the FRAILESEL data dictionary (see Table 1 for further information). The Study Director and the Study Coordinator will then identify unacceptable data entries using custom Excel queries to detect missing, impossible and improbable values and logical inconsistencies between data fields and across the forms. The DCC will then ask the sites to check for the incomplete data, and once the sites have resolved the data queries, the DCC will update the patient records. To identify complications during follow-ups, each center will check their database to monitor visits to the emergency department, postoperative imaging or intervention, outpatient visits or hospital readmissions. Additional checking of admission diagnosis and surgical procedures in the study months will identify any missing patients. The data will be collected from each individual center according to the current Italian Law regarding privacy policy (Legislative Decree no. 196/2003 "RIGHT TO PERSONAL DATA PROTECTION CODE"). It will be the responsibility of the local investigators to ensure that the local data will be protected and held according to such privacy policy and in line with what has been approved by the ethics board.

No patients are involved in setting the research question or the outcome measures; nor are they involved in the design and implementation of the study. There are no plans to involve patients in dissemination of results.

The following timeline has been outlined, to define specific stages of the study.

The report of this study will be prepared in accordance with guidelines set by the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement for observational studies ³⁰ . Statistical analysis will be performed with SPSS software, version 21 (IBM Analytics Italy, Segrate, MI) for MacOSX and/or MedCalc version 14.10.2 (MedCalc Software, Ostend, Belgium) for Windows. Based on previous reports with a comparable case mix, the expected 30-day mortality rate of the population in this study was estimated to be 20%. First, data normality will be tested using the Shapiro-Wilk test or Kolmogorov-Smirnov test. Dichotomous data and counts will be presented in frequencies. Continuous data will be presented as mean values plus standard deviations, or as median values and interquartile ranges. The 95% confidence interval will always be reported where appropriate. Differences between means will be compared using the independent sample Student t-test, the pairwise comparison Student t-test, the Mann-Whitney U test, the Kruskal-Wallis test or other analysis of variance (ANOVA) tests. To compare differences in frequencies, Fisher's exact test or χ ² test, with or without Yates’ correction will be performed. Receiver Operating Characteristic (ROC) curve analysis will be performed to estimate sensitivity and specificity of each score. Linear correlation will be assessed by Pearson’s or Spearman’s test, if needed. Multivariate analyses will be performed using linear or logistic regression models that consider mortality and morbidity as dependent variables. A p-value of <0.05 will be considered statistically significant.

The results of the FRAILESEL study will be disseminated through national and international conference presentations and peer-reviewed journals. The results will also be available through the study record website at ClinicalTrials.gov. Furthermore, additional studies and publications could be performed that analyse specific aspects of the data that will be presented. We are committed to ensuring that appropriate recognition is given to everyone who works on the study.