Murias et al. Critical Care (2016) 20:258 DOI 10.1186/s13054-016-1436-9 RESEARCH Open Access Automatic detection of ventilatory modes during invasive mechanical ventilation Gastón Murias1, Jaume Montanyà2,3, Encarna Chacón3, Anna Estruga3, Carles Subirà4, Rafael Fernández2,4, Bernat Sales5, Candelaria de Haro3, Josefina López-Aguilar2,3, Umberto Lucangelo6, Jesús Villar2,7, Robert M. Kacmarek8 and Lluís Blanch2,3* Abstract Background: Expert systems can help alleviate problems related to the shortage of human resources in critical care, offering expert advice in complex situations. Expert systems use contextual information to provide advice to staff. In mechanical ventilation, it is crucial for an expert system to be able to determine the ventilatory mode in use. Different manufacturers have assigned different names to similar or even identical ventilatory modes so an expert system should be able to detect the ventilatory mode. The aim of this study is to evaluate the accuracy of an algorithm to detect the ventilatory mode in use. Methods: We compared the results of a two-step algorithm designed to identify seven ventilatory modes. The algorithm was built into a software platform (BetterCare® system, Better Care SL; Barcelona, Spain) that acquires ventilatory signals through the data port of mechanical ventilators. The sample analyzed compared data from consecutive adult patients who underwent >24 h of mechanical ventilation in intensive care units (ICUs) at two hospitals. We used Cohen’s kappa statistics to analyze the agreement between the results obtained with the algorithm and those recorded by ICU staff. Results: We analyzed 486 records from 73 patients. The algorithm correctly labeled the ventilatory mode in 433 (89 %). We found an unweighted Cohen’s kappa index of 84.5 % [CI (95 %) = (80.5 %: 88.4 %)]. Conclusions: The computerized algorithm can reliably identify ventilatory mode. Keywords: Mechanical ventilation, Automatic detection, Ventilatory mode, Information systems in critical care Background underexploited for two main reasons. First, many data Monitoring is one of the main reasons for admission to are typically presented only fleetingly on screens that cli- intensive care units (ICUs). Up to 77 % of admissions to nicians see only when they are at the patient’s bedside. medical ICUs take place, at least in part, for monitoring So, unless an alarm is triggered, hours of acquired data purposes, even though only 10 % of the patients only are lost [2]. Trended data is of little use in identifying monitored will subsequently have indications for major asynchronies since asynchronies can occur sporadically interventions [1]. Accordingly, huge investments in and need to be identified as they occur, and displayed as monitoring technology have led to the development of a occurring at a frequency over time. Trending data is wide array of monitoring devices (bedside monitors, most useful in identifying physiologic change not short- mechanical ventilators, special devices, etc.) that gener- term events. Second, data that is not interpreted are use- ate large quantities of data. However, these data are less; to become useful, data must become information, by being processed, organized, structured, and contextu- * Correspondence: lblanch@tauli.cat alized [3]. Valuable information has often remained bur- 2 CIBER de Enfermedades Respiratorias, Instituto de Salud Carlos III, Madrid, ied even when the data in which it is based were widely Spain 3 Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí available. For instance, invasive monitoring of arterial (I3PT), Universitat Autònoma de Barcelona, c. ParcTaulí 1, 08208 Sabadell, blood pressure has been used since the beginnings of Spain ICUs, but in recent years the analysis and processing of Full list of author information is available at the end of the article © 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Murias et al. Critical Care (2016) 20:258 Page 2 of 7 these data has yielded information far more valuable Setting than arterial blood pressure alone. For example, pulse The study took place in two general ICUs (Parc Taulí contour can be used to estimate cardiac output [4, 5], University Hospital, Sabadell, Spain and Hospital Sant and pulse pressure variation can predict the cardiovascu- Joan de Deu-Fundació Althaia, Manresa, Spain) with 18 lar system’s response to a fluid load [6–8]. Tools to beds (4 beds in one hospital and 14 in the other) transform data from mechanical ventilators into mean- equipped with the Better Care® platform and one of the ingful information should help critical care clinicians an- following ventilators: Evita 4 (Dräger, Lübeck, Germany), ticipate harmful events, and expert systems could help Puritan Bennet 840 (Covidien, Plymouth, MN, USA), or them solve problems quickly and obtain expert advice. Servo I (Maquet, Fairfield, NJ, Sweden). The insti- The Better Care® software platform (Better Care S.L., tutional review board of the Parc Tauli University Barcelona, Spain) processes (standardizes, resamples, Hospital approved the protocol and waived informed synchronizes, analyzes, and stores) data from the data consent because the study was non-interventional, ports of mechanical ventilators or other monitoring de- posed no added risk to the patient, and did not inter- vices [9]. It has a set of algorithms that evaluate the be- fere with usual care. havior of airway pressure and flow to automatically detect potentially harmful events (ineffective inspiratory Patients efforts, double triggering, airway secretions, short and We studied consecutive patients aged >18 years admit- long cycles, aborted inspirations). In a previous study ted to one of the equipped beds who underwent mech- assessing the prevalence and time course of asynchronies anical ventilation for >24 hours. throughout mechanical ventilation, we found a median of 3.41 % asynchronies in all patients [10]. However, that study probably underestimated the magnitude of the Protocol problem because many of the analyses were limited to A team composed of two nurses and one physician the expiratory part of the flow tracing, where the shape who were not involved in any clinical decisions re- of the curve does not depend on the ventilatory mode. corded the ventilatory mode at the bedside once a Most device communication protocols do not provide in- day; the resulting log (including a timestamp for each formation about the ventilatory mode in use, and different entry) was used as a gold standard against which the manufacturers use different, often meaningless, names for Better Care® system’s automatic assignments were identical modes [11, 12]. For a system to detect events compared. during inspiration, it must be able to identify the ventila- The Better Care® software recorded and analyzed air- tory mode. A system that could identify the ventilatory way pressure and flow waveforms and calculated tidal mode would also enable protocols to facilitate communi- volumes for every breath during the hour preceding the cation between medical devices and expert systems by team’s recording of ventilatory mode and applied a two- converting data from proprietary device languages to stan- step algorithm to determine the specific ventilatory dardized formats guaranteeing interoperability [13]. The mode. The first step analyzes each breath and classifies aim of the present study is to validate an algorithm to it in one of seven categories according to the behavior of identify the most prevalent ventilatory modes based on inspiratory time (TI)s, flow, airway pressure (PAW), and the analysis of airflow and airway pressure waveforms. tidal volume (VT) (Table 1). To assess the stability of Methods Table 1 Breath classification strategy Software Breath characteristics The Better Care® system interacts directly with the signal output of medical devices through a device-specific connec- TI Flow Flow Slp PAW PAW lev Volume P300 tion driver. Mechanical ventilators and bedside monitors Type1 V V V V 1 V No are connected to the system using an ED41000P2-01 Type2 C C C V 2 C No remote access server (Lantronix, Irvine, CA, USA). The Type3 C V C V 2 C No system standardizes the signals, associating each recorded Type4 C V V C 2 V No curve with the parameter it represents, translates propri- Type5 V V V C 2 V No etary names into standard names, and resamples signals to Type6 V V V V 2 V Yes a frequency of 200Hz. Standardized signals are then analyzed, tagged, converted to Digital Imaging and Com- Type7 V V V V 2 V No munication in Medicine (DICOM) format, and stored in Abbreviations: TI inspiratory time, Flow inspiratory flow, Flow Slp inspiratory flow slope, PAW peek airway pressure, PAW lev number of PAW levels, Volume the hospital’s picture archiving and communication system tidal volume, P300 300 msec pause between inspiratory an expiratory time, (PACS). C constant, V variable Murias et al. Critical Care (2016) 20:258 Page 3 of 7 the different parameters, the system calculates a variabil- Type 7 breaths are present); spontaneous proportional ity index (VI) as follows: assist or proportional assist ventilation + (PC-CSVR+) (at sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi least 80 % of breaths are classified as Type 6 or Type 7). ðActual Value − Mean ValueÞ2 If the proportion of breaths classified did not fall into VI ¼  100 one of the above categories, the system labeled the rec- Mean Value ord as other modes. This label encompassed modes be- where “Actual Value” is the measured value for a given yond the scope of this algorithm, records that the variable and “Mean Value” is the running mean value for system was unable to correctly classify, and hours in the same variable over the last 20 breaths (even though which the ventilatory mode changed. step 1 of the algorithm is breath-based, the stability of a given parameter is evaluated in the context of the pre- Statistical analysis ceding breaths). In other words, VI represents the vari- We used the unweighted Cohen’s kappa coefficient to ation in the parameter as a percentage of the mean value assess the agreement between the team’s recordings in for the last 20 breaths. A variable was considered to be the log (gold standard) and the Better Care® mode detec- constant if the VI was less than 10 %. tion algorithm. The second step identifies the ventilator mode in pro- portion of the breaths classified into each category in Results the hour being analyzed (Table 2): continuous positive The team recorded the ventilatory mode 486 times airway pressure (CPAP) (at least 90 % of breaths are (Sabadell: n = 301; Manresa: n = 185) in 73 patients classified as Type 1); volume control-continuous (Sabadell: n = 31; Manresa: n = 42). The algorithm cor- mandatory ventilation or volume-controlled ventilation rectly labeled 433 (89 %) hours of ventilation mode with constant flow (VC-CMV) (at least 90 % of breaths (Table 2). The unweighted Cohen’s kappa coefficient was are classified as Type 2: volume control-continuous 84.5 % [95 % CI: 80.5–88.4 %)]. mandatory ventilation with decelerating flow or volume The system labeled 56 (12 %) records as other modes; control ventilation with decelerated flow (VC-CMVDF) this label was correct in 23 cases (5 % of the total) be- (at least 90 % of breaths are classified as Type 3); cause the mode did not fit any of the studied modes. pressure control-continuous mandatory ventilation or The system mislabeled a total of 53 (11 %) records; of pressure-controlled ventilation (PC-CMV) (at least 90 % these 20 (4 % of the total) corresponded to recordings of breaths are classified as Type 4); pressure control- mislabeled as one of the predefined modes and 33 (7 % continuous spontaneous ventilation or pressure support of the total) corresponded to recordings mislabeled as ventilation (PC-CSV) (at least 90 % of breaths are classi- other modes. Thus, the BetterCare® System was able to fied as Type 5); spontaneous proportional assist or pro- correctly label the mode 89 % of the time. portional assist ventilation (PC-CSVR) (which also includes neurally adjusted ventilatory assist, NAVA) (at Discussion least 80 % of breaths are classified as Type 6 and no Successful clinical decision support systems must pro- vide patient-specific recommendations [14]. To do that, Table 2 Mode classification strategy the system requires unequivocal information, at least for Breath type some critical variables. For systems advising clinicians 1 2 3 4 5 6 7 about mechanical ventilation, the ventilatory mode being CPAP >80 % used is critical information that current ventilator com- VC-CMV >80 % munication protocols do not readily provide. The algo- rithm incorporated in the BetterCare® system correctly VC-CMVDF >80 % identified the ventilatory mode in 433 (89 %) of the PC-CMV >80 % 486 hours recorded in 73 patients, with a kappa index of PC-CSV >80 % 84.5 % [95 % CI: 80.5–88.4 %]. Landis and Koch [15] PC-CSVR+ >90 % proposed the following standards for strength of agree- PC-CSVR 0% >90 % ment for the kappa coefficient: ≤0 % = poor, 1–20 % = Abbreviations: CPAP continuous positive airway pressure, VC-CMV volume slight, 21–40 % = fair, 41–60 % = moderate, 61–80 % = control-continuous mandatory ventilation or volume-controlled ventilation substantial and 81–100 % = almost perfect. Although a with constant flow, VC-CMVDF volume control-continuous mandatory ventilation kappa index of 84.5 % is highly acceptable we still mis- with decelerated flow; or volume-controlled ventilation with decelerated flow, PC-CMV pressure control-continuous mandatory ventilation or pressure-controlled labeled 11 % of the cases and continue to improve our ventilation, PC-CSV pressure control-continuous spontaneous ventilation or recognition algorithms. pressure support ventilation, PC-CSVR spontaneous proportional assist or proportional assist ventilation, PC-CSVR+ spontaneous proportional assist or Expert systems dealing with mechanical ventilation need proportional assist ventilation + to be able to detect ventilatory modes automatically. The Murias et al. Critical Care (2016) 20:258 Page 4 of 7 number of ventilatory modes has increased dramatically four from problems in differentiating between PC-CMV over the last 30 years. Moreover, within each mode many and PC-CSV. Another six records where the actual features can be activated that transform the basic mode mode was VC-intermittent mandatory ventilation with into what is effectively another mode. To add further con- pressure support were labeled PC-CSV (two cases) or fusion, the lack of standards in ventilatory mode naming PC-CMV (four cases). [11, 12] has resulted not only in different names for the A critical aspect of our algorithms is to establish if a same ventilatory mode but also in different ventilatory given parameter is (or is not) stable. If a parameter modes with the same name. Although manufacturers’ changed more than 10 %, the algorithm considers it names can be translated into standard names, most com- “variable”. To reduce the threshold below 10 % would munication protocols from mechanical ventilators do not reduce the number of variable parameters considered provide information on the ventilatory mode in use. We constant. Unfortunately, it would also increase the num- chose these modes because they accounted for 70 % of the ber of constant parameters considered variable: even ventilatory time in a recent large international study on when these thresholds typically are more precise than the epidemiology of mechanical ventilation [16], and cor- required by the ISO standard [18], for delivery of pres- responded to 94 % of the modes used in our cohort. sures, flows, and volumes by ventilators. The practical consequences of failing to label a record For instance, differentiating PC-CMV from PC-CSV is are different from those of mislabeling a record. If the based on the variability of inspiratory time and distin- system cannot determine the ventilatory mode, it cannot guishing between VC-CMVDF and PC-CMV, which re- provide advice; however, if the system mislabels the lies on the variability of airway pressure, the slope of the mode, it may provide erroneous advice. For example, air- inspiratory flow curve and tidal volume. However, a way pressure-time profile analysis could provide infor- fixed ventilatory pattern (or the absence of muscular ac- mation about two important determinants of ventilator- tivity) leads to a scenario in which patients ventilated in induced lung injury: tidal recruitment and overdistention PC-CMV will have a constant inspiratory volume and [17]. This analysis derives an index by fitting the central slope of inspiratory flow and patients in VC-CMVDF will part of the inspiratory airway pressure-time profile to an have consistent airway pressure. Our analysis of mis- exponential function. Importantly, however, the index is labeled records showed that most mistakes occurred in valid only if the ventilatory mode is VC-CMV and no pa- records with very regular ventilatory patterns where the tient inspiratory efforts are present [17]. Unless both variation in parameters expected to vary was below the these conditions are met, there is a danger of misinter- tolerance limits we had set. Furthermore, the high in- pretation that could lead to incorrect decisions and place spiratory pressure alarm, for instance, can abort inspir- the patient at risk. In the present study, 20 (4 %) records ation making parameters expected to be constant were mislabeled. Of these, eight resulted from problems become very variable (like inspiratory time in PC-CMV in differentiating between PC-CMV and VC-CMVDF and and VC-CMV or tidal volume in VC-CMV). Fig. 1 Irregular traces from a patient in PC-CMV. A second inspiratory peak flow in a machine-triggered breath strongly suggests reverse triggering Murias et al. Critical Care (2016) 20:258 Page 5 of 7 Fig. 2 The airway pressure alarm aborts inspiratory cycles, producing highly variable inspiratory times and volumes and making it difficult for the system to correctly identify the mode as VC-CMV A total of 56 (12 %) records were not assigned to records that were difficult even for expert physicians one of the modes. In 23 cases the decision was cor- to classify (Figs. 1 and 2). Another problem, specific rect, because the actual mode was one of the modes to Servo ventilators, is where in VC-CMV the inspira- that were outside the scope of the algorithm. In 33 tory valve opens to mitigate flow asynchrony, causing cases, however, the system should have assigned the tidal volume to become highly variable and causing mode. The main reason for these mistakes was severe the system to fail to recognize the mode as VC-CMV patient-ventilator asynchrony that resulted in muddled (Fig. 3). Fig. 3 Manufacturer’s non-standard implementation of ventilatory modes. With the goal to increase patient comfort and reduce flow asynchrony, some manufacturers have developed modifications of classic ventilatory modes. For instance, Maquet ventilators in VC-CMV (when airway pressure shows a concavity, inadequate flow to meet patient inspiratory demand), the ventilator opens its demand valve allowing additional gas flow to avoid flow asynchrony leading to an increase in tidal volume potentially violating a lung-protective approach. Beyond its advantages and drawbacks, it causes changes in the inspiratory flow profile and tidal volume that prevent the system from identifying the mode as VC-CMV Murias et al. Critical Care (2016) 20:258 Page 6 of 7 Determining the ventilator mode is also very Acknowledgements important for detecting the occurrence of inspiratory The authors thank Mercè Ruiz and Mr. John Giba for their invaluable support in editing the manuscript. patient-ventilator asynchrony (flow asynchrony, delayed triggering, and patient-ventilator inspiratory time mis- Funding match). In a recent study of patient-ventilator asyn- This work was funded by projects PI09/91074 and PI13/02204, integrated in chronies in a general population monitoring the entire the Plan Nacional de R+D+I and co-funded by the ISCIII- Subdirección Gen- eral de Evaluación y el Fondo Europeo de Desarrollo Regional (FEDER). CIBER period of mechanical ventilation, Blanch et al. [10] found Enfermedades Respiratorias, Fundación Mapfre, Fundació Parc Taulí, Plan that asynchronies are very frequent and are associated Avanza TSI-020302-2008-38, MCYIN and MITYC (Spain). with mortality. However, they probably underestimated Role of the sponsors: no funding organization or sponsor was involved in the design or conduct of the study, in the collection, management, analysis, the prevalence of ineffective inspiratory efforts because or interpretation of the data, or in the preparation, review, or approval of the the inability to determine the ventilatory mode limited manuscript. the system’s analyses mostly to the expiratory portion of the ventilatory cycle [9]. Incorporating an algorithm to Availability of data and materials The database on which this study is based is available as a spreadsheet in detect the ventilatory mode in use would enable the en- Additional file 1. tire ventilatory cycle to be analyzed and provide a more complete picture of the problem. Authors’ contributions This study’s main strength is that it represents real LB had full access to all of the data in the study and takes responsibility for problems occurring in ICUs. The records analyzed came the integrity of the data and the accuracy of the data analysis. LB, JM, UL, GM, RF, JV, and RMK contributed to the study concept and design. BS, JM, from a wide variety of real ICU patients with asyn- EC, and AE carried out acquisition of data. LB, JM, UL, EC, AE, JL-A, CS, RF, JV, chronies, cough, airway secretions, etc. Its main limita- GM, and RMK performed analysis and interpretation of data. LB and GM tion is that it considered a limited set of ventilatory drafted the manuscript. LB, BS, JM, UL, EC, AE, JL-A, RF, JV, GM, and RMK critically revised the manuscript for important intellectual content. JM and modes. We have not developed algorithms to detect dual GM performed the statistical analysis. LB, BS, and JM provided administrative, modes because their use remains marginal [16]. More- technical, or material support. LB, GM, JV, and RMK conducted the study over, the low number of proportional pressure support supervision. All authors read and approved the final manuscript. records in our database precludes the validation of an al- Competing interests gorithm to detect those modes. Thus, the reliability of LB, BS, and GM are inventors of a US patent owned by Corporació Sanitaria our algorithms to detect ventilatory modes has been Parc Taulí: “Method and system for managing related patient parameters only established for VC-CMV, VC-CMVDF, PC-CMV, provided by a monitoring device,” US Patent No. 12/538,940. LB, BS, GM, and UL own stock options in BetterCare S.L., a research and development spinoff PC-CSVs PC-CSVR+, and CPAP. of Corporació Sanitària Parc Taulí (Spain). RMK is a consultant for Covidien and OrangeMed, has received research grants from Covidien and Venner Medical. JV has received research grants from Maquet. JM, EC, AE, CS, CH, Conclusions JL-A, and RF have no competing interests. Automatic systems can accurately identify the most Consent for publication commonly used ventilatory modes, thus providing cru- Not applicable. cial information that enables the entire ventilatory cycle to be analyzed. More data must be collected to validate Ethics approval and consent to participate the algorithm for identifying less frequently used propor- The ethics committees at both institutions approved the study and decided tional pressure support modes. to waive informed consent because the study was observational, signals were encrypted to ensure privacy, no extra effort to on-charge personnel was generated, no changes to usual treatment were required, and no information that could lead to changes in patient care was provided to the Additional file attending team. Author details Additional file 1: The database on which this study is based. (XLS 77 kb) 1 Clínica Bazterrica y Clínica Santa Isabel, Departamento de Ciencias Fisiológicas, Farmacológicas y Bioquímicas, Facultad de Medicina, Universidad Favaloro, Buenos Aires, Argentina. 2CIBER de Enfermedades Abbreviations Respiratorias, Instituto de Salud Carlos III, Madrid, Spain. 3Parc Taulí Hospital CPAP, continuous positive airway pressure; DICOM, Digital Imaging and Universitari, Institut d’Investigació i Innovació Parc Taulí (I3PT), Universitat Communication in Medicine; ICU, intensive care unit; NAVA, neurally Autònoma de Barcelona, c. ParcTaulí 1, 08208 Sabadell, Spain. 4Servei de adjusted ventilatory assist; PACS, picture archiving and communication Medicina Intensiva, Hospital Sant Joan de Deu-Fundació Althaia, Universitat system; PAW, airways pressure; PC-CMV, pressure control-continuous Internacional de Catalunya, Manresa, Spain. 5Better Care, Barcelona, Spain. 6 mandatory ventilation or pressure-controlled ventilation; PC-CSV, pressure Department of Perioperative Medicine, Intensive Care and Emergency, control-continuous spontaneous ventilation or pressure support ventilation; Cattinara Hospital, Trieste University, Trieste, Italy. 7Research Unit, Hospital PC-CSVR, spontaneous proportional assist or proportional assist ventilation; Universitario Dr. Negrín, Las Palmas de Gran Canaria, Spain. 8Department of PC-CSVR+, spontaneous proportional assist or proportional assist ventilation +; Respiratory Care, Massachusetts General Hospital and Department of TI, inspiratory time; VC-CMV, volume control-continuous mandatory ventilation Anesthesiology, Harvard Medical School, Boston, MA, USA. or volume-controlled ventilation with constant flow; VC-CMVDF, volume control- continuous mandatory ventilation with decelerated flow; or volume-controlled Received: 12 April 2016 Accepted: 22 July 2016 ventilation with decelerated flow; VI, variability index; VT, tidal volume Murias et al. Critical Care (2016) 20:258 Page 7 of 7 References 1. Thibault GE, Mulley AG, Barnett GO, Goldstein RL, Reder VA, Sherman EL, Skinner ER. Medical intensive care: indications, interventions, and outcomes. N Engl J Med. 1980;302(17):938–42. 2. Pickering BW, Gajic O, Ahmed A, Herasevich V, Keegan MT. Data utilization for medical decision making at the time of patient admission to ICU. Crit Care Med. 2013;41(6):1502–10. 3. Pinsky MR, Dubrawski A. Gleaning knowledge from data in the intensive care unit. Am J Respir Crit Care Med. 2014;190(6):606–10. 4. Alderman EL, Branzi A, Sanders W, Brown BW, Harrison DC. Evaluation of the pulse-contour method of determining stroke volume in man. Circulation. 1972;46(3):546–58. 5. Wesseling KH, Jansen JR, Settels JJ, Schreuder JJ. Computation of aortic flow from pressure in humans using a nonlinear, three-element model. J Appl Physiol (1985). 1993;74(5):2566–73. 6. Auler Jr JO, Galas F, Hajjar L, Santos L, Carvalho T, Michard F. Online monitoring of pulse pressure variation to guide fluid therapy after cardiac surgery. Anesth Analg. 2008;106(4):1201–6. table of contents. 7. Berkenstadt H, Margalit N, Hadani M, Friedman Z, Segal E, Villa Y, Perel A. Stroke volume variation as a predictor of fluid responsiveness in patients undergoing brain surgery. Anesth Analg. 2001;92(4):984–9. 8. Hofer CK, Muller SM, Furrer L, Klaghofer R, Genoni M, Zollinger A. Stroke volume and pulse pressure variation for prediction of fluid responsiveness in patients undergoing off-pump coronary artery bypass grafting. Chest. 2005;128(2):848–54. 9. Blanch L, Sales B, Montanya J, Lucangelo U, Garcia-Esquirol O, Villagra A, Chacon E, Estruga A, Borelli M, Burgueno MJ, et al. Validation of the Better Care(R) system to detect ineffective efforts during expiration in mechanically ventilated patients: a pilot study. Intensive Care Med. 2012;38(5):772–80. 10. Blanch L, Villagra A, Sales B, Montanya J, Lucangelo U, Lujan M, Garcia- Esquirol O, Chacon E, Estruga A, Oliva JC, et al. Asynchronies during mechanical ventilation are associated with mortality. Intensive Care Med. 2015;41(4):633–41. 11. Chatburn RL, El-Khatib M, Mireles-Cabodevila E. A taxonomy for mechanical ventilation: 10 fundamental maxims. Respir Care. 2014;59(11):1747–63. 12. Chatburn RL, Volsko TA, Hazy J, Harris LN, Sanders S. Determining the basis for a taxonomy of mechanical ventilation. Respir Care. 2012;57(4):514–24. 13. Halpern NA. Innovative designs for the smart ICU: Part 3: Advanced ICU informatics. Chest. 2014;145(4):903–12. 14. Payne TH. Computer decision support systems. Chest. 2000;118(2 Suppl): 47S–52S. 15. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–74. 16. Esteban A, Frutos-Vivar F, Muriel A, Ferguson ND, Penuelas O, Abraira V, Raymondos K, Rios F, Nin N, Apezteguia C, et al. Evolution of mortality over time in patients receiving mechanical ventilation. Am J Respir Crit Care Med. 2013;188(2):220–30. 17. Grasso S, Terragni P, Mascia L, Fanelli V, Quintel M, Herrmann P, Hedenstierna G, Slutsky AS, Ranieri VM. Airway pressure-time curve profile (stress index) detects tidal recruitment/hyperinflation in experimental acute lung injury. Crit Care Med. 2004;32(4):1018–27. 18. ISO 80601-2-12: 2011/Cor 1: 2011. Medical electrical equipment - Part 2-12: Particular requirements for basic safety and essential performance of critical care ventilators. Geneva, Switzerland: International Organization for Standardization (ISO); 2011. Submit your next manuscript to BioMed Central and we will help you at every step: • We accept pre-submission inquiries • Our selector tool helps you to find the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit