Effect of Altitude on the Evolution of Acute Respiratory Distress Syndrome

HYPOTHESIS Physiological adaptations generated by living at high altitude modify the clinical evolution of ARDS. OBJECTIVE To describe the main clinical and developmental characteristics of ARDS diagnosed at altitudes greater than 1500 meters above sea level. To compare the main clinical and developmental characteristics of ARDS diagnosed at altitudes greater than 1500 meters above sea level with those of cases diagnosed at altitudes below 1500 meters above sea level. METHODOLOGY Main events: Incidence of ARDS Mortality of ARDS Duration of mechanical ventilation and length of ICU stay of patients with ARDS Sequelae found 1 year and 3 years after diagnosis of ARDS Secondary events Recognition of ARDS by clinician Modes of mechanical ventilation Use of adjuvant therapies Design: observational, prospective, cohort, multicenter, international study Recruitment period: July 18, 2.016 - August 18, 2.016 General Procedure The general organization chart of the study implies the participation of 3 lead researchers, at least 1 researcher/coordinator per country, and main researchers of each service. For the selection of participating services, a general invitation will be sent to all South American ICUs located at altitudes greater than 1500 meters above sea level, as well as a representative group of ICUs located below 1500 meters above sea level. The 2 cohorts (altitude and sea level) will be generated afterwards; the size of each cohort will be determined by convenience. For the selection of each participating service, the exploratory form (appendix 1) will be sent beforehand, in order to gain knowledge of its characteristics and possibilities; if the center meets the criteria, it will be enrolled and the rest of the documents (research protocols, appendices 2 and 3) will be sent to it. Once the participating services are defined, coordination videoconferences will be held before initiation of the study; an e-mail address will be set up for general coordination. All questions will be answered in less than 24 hours. The data will be collected in an Excel spreadsheet that will be sent to the coordinator of each country, and he will distribute it to the main researcher of each participating center. The information will be afterwards submitted to the general coordinators via the reverse pathway. All information will be sent by e-mail; in order to maintain confidentiality, the spreadsheets will be encrypted. The opening codes will be known only by the researchers responsible for each service, for each country, and by the lead researchers of the study. The data will be included by the main researcher of each center, under supervision of the coordinator of each country. Day 1 will be the day of admission to ICU with an unadjusted Pa02/Fi02 less than 300, regardless of the day the patient was admitted to hospital or to ICU. The protocol implies the gathering of basic data (appendix 3), that will be further obtained when the patients meets the criterion of Pa02/Fi02 less than 300. Gathering of data ends 28 days after day 1 (IT MAY NOT CORRESPOND TO DAY 28 AFTER ADMISSION TO ICU); afterwards, data such as date of extubation, of discharge from ICU and from hospital, and condition upon discharge from hospital will be gathered. Clinical and analytic data will be gathered on admission, and afterwards at the same hour of the morning, while the patient continues under invasive mechanical ventilation. Diagnosis of ARDS is made by two procedures: (A) automatic, which will consist of individually picking each variable pertaining to the definition, and afterwards, by way of a computer algorithm, defining whether the patient complies with the Berlin definition; (B) clinical, the attending physician will be asked daily, since the moment when PaO2/Fi02\<300, what the reason for hypoxemia is. Both diagnoses will be independent and double-blinded. The ratio of concordance will be calculated afterwards. Operational definitions: ARDS: Berlin definition. Days free of mechanical ventilation are the number of days in which no invasive mechanical ventilation is required, from disconnection to day 28. If the patient dies, 0 days free of mechanical ventilation will be considered. Driving pressure s the plateau pressure minus PEEP. Incidence is the number of patients diagnosed with ARDS divided by the number of patients who required invasive mechanical ventilation over the study period. Number of patients with ARDS per bed of the service is the number of patients diagnosed with ARDS divided by the number of beds suitable for mechanical ventilation available over the study period. Long-term follow-up (1 and 3 years) will be carried out by the central coordination team via e-mail or, occasionally, by telephone. Finally, a single database will be generated, with which the final report will be done. All patients with incomplete data will be excluded from the analysis. Statistical analysis: In order to fulfill objective 1, a descriptive epidemiological analysis of the primary and secondary events gathered in patients diagnosed with ARDS according to the Berlin definition. In order to fulfill objective 2, a comparison is made between the proportion of primary and secondary events between the altitude cohort and the sea-level cohort. Categorical variables are presented as number of cases and percentage, continuous variables are presented as median and percentiles 25-75. Categorical variables are compared by the chi-2 continuous variables are compared by Mann-Whitney or Kruskal-Wallis non-parametric tests. In order to fulfill objective 3, ROC curves will be made for each event, taking into consideration the unadjusted and altitude-adjusted Berlin definition as dependent variable. The final data report will be done following the recommendations of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement guidelines for observational cohort studies. The whole statistical analysis will be made with the free R program (http://www.R-project.org). All P values will be two-tail P values, considering a \<0.05 level of signification. ACCESSIBILITY OF DATA The data will be accessible to any researcher upon written request to the Latin American Network of High-Altitude Intensive Care Medicine (RELAMI). The request must be written in Spanish and include: justification for using the data commitment to using the data for exclusively scientific purposes (explicitly stating that they will not be used for economic purposes) commitment to publishing the data in a pubmed indexed journal within a one-year period proof of experience and background in the field of clinical research adhesion to the code of ethics RELAMI will analyze the request and decide within a 60-day period from receipt of the request. If the response is positive, the data will be sent anonymized; if it is negative, the reasons will be explained. ECONOMIC ASPECTS This study does not have sponsors. No researcher will receive economic compensation or any other type of reward for participating. ETHICAL ASPECTS This study will be carried out in accordance with the ethical principles established by the International Conference on Harmonization, guidelines for Good Clinical Practice (GCP), which stem from the Helsinki Declaration and applicable regulation norms. This study is considered to be of insignificant low risk, since it is an observational, non-interventional study, and the data are gathered without identification. Each participating ICU will process the pertaining institutional approvals, including a renunciation of ethical revision. Recorded dispensation of each participant for gathering clinical data without identification will be sought as part of routine clinical care. DATA MANAGEMENT AND DISSEMINATION OF RESULTS The lead researchers and the Coordination Center will act as guardians of the data. Reports of progress and final reports will be presented opportunely in national and international conferences, giving priority to presentation during annual conferences in Latin America. The results of the study will be published in indexed scientific journals, preferably of the first decile. Only aggregated and disidentified data will be publicly presented.