Introduction

The increasing clinical use of critical care echocardiography (CCE) is paralleled by the growing need for scientific knowledge in the field [1, 2]. PRICES, an acronym standing for “Preferred Reporting Items for Critical care Echocardiography Studies”, is a project endorsed by the European Society of Intensive Care Medicine (ESICM) and conducted by the Echocardiography Working Group of the Cardiovascular Dynamic section. Its final aim is to improve methodological and reporting consistencies in clinical research in the areas of left venticular (LV) systolic function, LV diastolic function, right ventricular (RV) systolic function and fluid management.

Two critical methodological aspects for CCE research influence the knowledge in the field: (1) a well-structured approach to the single CCE exam with detailed information provided together with a well-described clinical context; (2) a detailed description of methodology and reporting of results, producing comparable echocardiography data between research studies.

The present article contains recommendations provided by the panel of echocardiography experts involved in PRICES. Notably, the PRICES recommendations are based not only on the experts’ opinion on the importance of each item that may warrant reporting, but also on an extensive systematic review and literature appraisal conducted by the authors on the currently published CCE literature [3]. Such appraisal provided the panel with more objective insights on the necessity of improving reporting of particular items in echocardiography research. It is desirable that the list of recommendations of PRICES will work as “checklists” for authors as to which parameters and information could/should be included in the design of their echocardiography research studies and subsequently reported in their manuscript on CCE. However, it is important to highlight that recommendations on how to conduct the CCE at bedside and the echocardiographic measurements are beyond the scope of PRICES as many valuable international guidelines provide ample direction on these subjects.

Methods

In brief, the PRICES project was initiated by the Echocardiography Working Group of the ESICM and started with a selection of 19 experts in the field of CCE, from Europe (n = 15), Oceania (n = 3) and North America (n = 1); the first internal discussion of the PRICES group was held in Vienna (25th–26th September 2017). The experts agreed on a list of items that are of potential interest in CCE research studies. Subsequently, these items were appraised with a systematic approach described in the PRICES part I. The systematic review was registered on PROSPERO (CRD42018094450). The literature searches were performed separately for each topic/area on Medline and Embase including studies published from 1st January 2000 to 31st December 2017, as reported in the PRICES part I. Two experts screened each abstract retrieved from the search, and included articles were appraised for a list of pre-determined items [3]. A total of 43 “common items” were defined as items of interest to all CCE studies, and a variable number of “topic-specific items” according to 5 different areas of CCE interest. Specifically, 15 “topic-specific items” were selected for studies involving the evaluation of left ventricular systolic function (LVSF), 18 for right ventricular function (RVF), 15 for LV diastolic function (LVDF), 7 for fluid management (FM) and 17 for advanced echocardiography techniques (AET, including speckle tracking and 3D echocardiography). However, although the area in AET was initially planned for recommendations and had been reported in the systematic review [3], the panel decided that AET is currently at early stage of its introduction into critical care practice and thus it is premature to give formal recommendations. Instead, recommendations as to the use of AET, such as strain measurements, were left to each area (e.g. LV and RV systolic functions) to deliberate.

Evaluation of the importance of items

The process from starting the project to finally establishing the guidelines is summarized in Fig. 1. Importantly, since each panel expert might opine the importance for any item differently, the experts agreed to rate the relative importance (RI) for each item independently. The RI is a measure, on a Likert scale from 1 to 4 points, of the opined importance of an item. The final overall RI for each item was determined on majority votes. Items were thus dived in “essential” (4 points), “additional” (3 points), “optional” (2 points) and “not important” (1 point). The RI was scored on the basis of clinical utility or likelihood of biased and non-reproducible results if that item was not reported; hence the higher the RI, the greater the propensity for misinterpretation if missing.

Fig. 1
figure 1

Processes leading up to the final recommendations. The topics of studies and items were specified in the first expert panel meeting. Each item was given due consideration and the relative importance (RI) for each item was rated by the experts. These items were divided into “core” or “supplementary” items based on RI. The percentage of reporting for each item was calculated from a body of evidence (research publications) for each topic. The “core items” were then subdivided into those that need or do not need high priority discussion based on the popularity (percentage) of reporting. These items were those which deemed “important” (RI = 3 or 4) by the experts but had a low reporting rate. The panel then decides which items were to be “Essential” or “Additional” after in-depth discussions. The decisions were based on usefulness, utility, clinical perspective and feasibility

Full size image

The “essential” (RI = 4) and “additional” (RI = 3) reflected “core items” in the interpretation of CCE research findings; on the contrary, items with RI score of 1 or 2 were considered “supplementary”, and the decision of reporting them can be left to the authors of scientific studies. This represented a first step based only on the panel’s expertise and experience in CCE.

Combining item importance with the frequency of reporting

In a subsequent step, we borrowed the principle of Hand rule (burden of risk calculation) [4] to use a metric that we named “critical index” (CI) with the aim of balancing the importance of each item with its frequency of reporting (FSi), as reported in our systematic literature appraisal on CCE [3]. In brief, the Hand rule provides that the burden (cost) of risk prevention (B) is the product of the probability of events (P) and the gravity of loss if the event happened (L): B = P × L. Applying this principle, the CI can be seen as the burden of prevention of not reporting an item, and can be expressed as

$$\mathrm{CI}=\left(1-\mathrm{FSi}\right)\times \mathrm{RI},$$

where (1 − FSi) was the fraction of studies that did not report the item and is equivalent to the probability of missing an item, and RI can be read as the impact (gravity) of an item if not reported. Hence, a low CI indicates lesser burden, and high CI indicates larger burden and some “precautions” are warranted. These “precautions” were stated in the form of whether or not an item should be reported more often in this recommendation. When considering the need to increase reporting of an item, we arbitrarily divided the CI values in three categories:

  • CI ≤ 1: the scientific community is already aware of the importance of reporting the item and/or the RI is low hence the burden to recommend is low (green flag).

  • 1 < CI < 3, increased reporting is needed (orange flag).

  • CI ≥ 3, the scientific community is not aware of the importance of reporting the item and/or the RI is high, hence the burden to recommend is high (red flag).

The advantage of this approach is the combination of an objective measure (1 − FSi) with the experts’ opinions (RI). It is the opinion of the panel that the CI gives a balanced idea of the need to increase reporting of each item.

The rationale for the recommendations is discussed in the text, and tables are also provided with summaries of the items. An easy-to-follow PRICES utility checklist is made available to readers (researchers and clinicians) in Fig. 2.

Fig. 2
figure 2figure 2

PRICES utility checklist

The checklist is provided to assist clinicians and researchers to collect information that we regard as useful and important. We divided the information into several domains, some of which are important in any echo study, others are specific to the aim of the study. There are altogether 5 Tables for data collection. Table A, items that are common to all topics, Table B, items that are deemed important or useful in reporting LV systolic function, Table C, items that are deemed important or useful in reporting RV systolic function, Table D, items that are deemed important or useful in reporting LV diastolic function, Table E, items that are deemed important or useful in studies related to fluid management

This checklist can also be used as data collection form. Checkboxes that are not shaded indicate essential item that should be collected and reported. Shaded checkboxes indicate additional information that are useful for interpretation

Full size image

The recommendations

Items common to all topics

Supplementary Material eTable 1 shows the list of the 43 common items divided in six sections, with different symbols according to the RI of each item with regard to CCE interest.

Study characteristics (3 items)

Study characteristics are all essential and well reported in the current literature. The panel reinforces the importance of stating if data are prospectively acquired, and to distinguish the number of exams from the number of patients (same patient may be exposed to more than one echocardiography exam).

Patients characteristics (12 items)

The clinical context should be always clear. In order to increase the reproducibility of the study results, it is essential to describe the baseline characteristics of the study population, in particular of the comorbidities. Knowledge of pre-existing cardiac dysfunction before critical illness is very valuable.

Echocardiography information (6 items)

An increase in the reporting of information on echocardiography exams is highly desirable. Such technical information is mostly judged as essential. Although optional in all fields of recommendations, software version is paramount when reporting AET parameters.

Clinical information at the time of echocardiography (10 items)

The overall reporting of data on ventilation (mode and settings) and on the hemodynamic conditions at the time of the echocardiography exam should increase. Information regarding mechanical ventilation is of greatest importance for the RVF and FM topic. Description of hemodynamic conditions and pharmacological support is essential for all sub-groups. It is of utmost importance to clearly report if patients were all in sinus rhythm, or whether patients with non-sinus rhythm or paced rhythm were excluded.

Measurement reliability (8 items)

Training of the echocardiographers/sonographers and of those who reviewed and interpreted the echocardiography exam are deemed essential to all CCE topics, though reporting was discontinuous across the CCE topics.

Statistics reporting (4 items)

Sample size calculation should be performed and deserves a significant increase in reporting. In the case of pilot studies, authors are encouraged to provide a reasonable estimation of sample size. It is important to state if analysis was blinded and if it addressed for potential confounding; it is of additional value to provide information on any internal validation of the study.

Topic-specific items

Figure 2 is a summary of the main recommendations and Figs. 3, 4, 5, and 6 report the recommendations in detail with items divided into parameters specifically describing the evaluation of the selected topic (LVSF, RVF, LVDF or FM) and those allowing a better understanding of the reported data, providing a better clinical picture.

Fig. 3
figure 3

Left ventricular (LV) systolic function recommendations

Items are divided into two groups: 1) those specifically evaluating the LV systolic function, and 2) in those allowing a better interpretation of the clinical context when the echocardiography exam is performed. The relative importance (RI) of the item is divided in those essential (E, RI score 4), additional (A, RI score 3) or optional (O, RI score 2). Essential and Additional items are coupled with a graded and coloured scale according to the necessity of increase attention in reporting by the scientific community, as gathered by the results of the critical index (CI). The grades are as follow: 3 - Significant Increase in reporting recommended CI ≥ 3 (red); 2 - Increase in reporting recommended 1.0 <CI<3 (orange); 1 - Continue reporting recommended CI ≤ 1.0 (green)

CCE: critical care echocardiography; FAC: fractional area change; LVEF: LV ejection fraction; MAPSE: mitral annular plane systolic excursion; RWMAs: regional wall motion abnormalities; TDI: tissue Doppler imaging; VTI: velocity time integral

Full size image
Fig. 4
figure 4

Right ventricular (RV) function recommendations

Items are divided into two groups: 1) those specifically evaluating the RV function, and 2) in those allowing a better interpretation of the clinical context when the echocardiography exam is performed. The relative importance (RI) of the item is divided in those essential (E, RI score 4), additional (A, RI score 3) or optional (O, RI score 2). Essential and Additional items are coupled with a graded and coloured scale according to the necessity of increase attention in reporting by the scientific community, as gathered by the results of the critical index (CI) The grades are as follow: 3 - Significant Increase in reporting recommended CI ≥ 3 (red); 2 - Increase in reporting recommended 1.0 <CI<3 (orange); 1 - Continue reporting recommended CI ≤ 1.0 (green)

CCE: critical care echocardiography; EDA: end-diastolic area; EDD: end-diastolic diameter; FAC: fractional area change; IAS: inter-atrial septum; IVC: inferior vena cava; LV: left ventricle; PAAT: pulmonary artery acceleration time; PAPs: pulmonary artery systolic pressure; TAPSE: tricuspid annular plane systolic excursion; TDI: tissue Doppler imaging; TR: tricuspid regurgitation

Full size image
Fig. 5
figure 5

Left ventricle (LV) diastolic function recommendations

Items are divided into two groups: 1) those specifically evaluating the LV diastolic function, and 2) in those allowing a better interpretation of the clinical context when the echocardiography exam is performed. The relative importance (RI) of the item is divided in those essential (E, RI score 4), additional (A, RI score 3) or optional (O, RI score 2). Essential and Additional items are coupled with a graded and coloured scale according to the necessity of increase attention in reporting by the scientific community, as gathered by the results of the critical index (CI). The grades are as follow: 3 - Significant Increase in reporting recommended CI ≥ 3 (red); 2 - Increase in reporting recommended 1.0 <CI<3 (orange); 1 - Continue reporting recommended CI ≤ 1.0 (green)

PAPs: pulmonary artery systolic pressure; TDI: tissue Doppler imaging; TR: tricuspid regurgitation.

Full size image
Fig. 6
figure 6

Fluid management recommendations

Items are divided into two groups: 1) those specifically describing the methods of preload variation and of fluid responsiveness (FR) assessment, and 2) in those allowing a better interpretation providing data on the reference standard methods and technical info. The relative importance (RI) of the item is divided in those essential (E, RI score 4), additional (A, RI score 3) or optional (O, RI score 2). Only Essential items were graded in this topic. Such items are coupled with a graded and coloured scale according to the necessity of increase attention in reporting by the scientific community, as gathered by the results of the critical index (CI). The grades are as follow: 3 - Significant Increase in reporting recommended CI ≥ 3 (red); 2 - Increase in reporting recommended 1.0 <CI<3 (orange); 1 - Continue reporting recommended CI ≤ 1.0 (green)

PLR: passive leg raising; SV, stroke volume; VC: volume challenge

Full size image

LV systolic function

We recommend (Fig. 3) more transparent reporting on the technical aspect for estimation of LV ejection fraction (LVEF) i.e. Simpson, Teicholz or 3D. Although LV size is essential for studies on LVSF, the panel does not feel it necessary to recommend one approach over another (diameter, area, volume). If feasible, studies focusing on LVSF are encouraged to integrate their information with data on the LVDF.

Additional parameters provide useful information regarding myocardial function (Mitral Annular Plane Systolic Excursion, LV S’ wave on Tissue Doppler Imaging (TDI), LV strain and strain rate) in the context of critical illness. For instance, the LV global longitudinal strain may be of better prognostic value when compared to LVEF in septic patients [5]. While the panel agrees that speckle tracking echocardiography is of great and increasing interest, most ICUs are probably not adequately equipped for strain analysis. For studies evaluating the LV with S’ wave on TDI or Mitral Annular Plane Systolic Excursion, it is also important to report regional wall motion abnormalities and the presence of mitral annular calcification/prosthesis.

Other parameters do not directly describe the LVSF but are essential for a correct interpretation of the research findings. We recommend reporting information on functional heart valve disease. An illustrative example of the presence of heart valve disease might be when mitral valve regurgitation causes significant overestimation of LVEF.

RV function

The assessment of RVF is challenging, in that a single accurate global measurement has not yet been identified. The varied nature of RV pathophysiology necessitates the use of a number of quantitative descriptors (Fig. 4). The majority of parameters were considered essential, including not only specific measurements of RVF but also items allowing accurate interpretation of RVF as evaluated with echocardiography. Three items describing RV contractility (RV fractional area change, tricuspid annular S’ with tissue Doppler imaging, tricuspid annular plan systolic excursion) were judged essential. In regard to tricuspid annular S’ wave, we recommend reporting the image plane and the location of the sampling point. Subjective RVF rating is optional and indeed a recent study confirmed that it should not be used in isolation [6].

Reporting RV size (including in comparison with the LV size) is a component of RVF [7, 8] and should be accompanied by the echocardiographic plane and the method employed. The reportage of pulmonary artery (PA) pressure is also important in interpreting RVF, and should be clear whether PA pressure is estimated from tricuspid regurgitation or derived from a PA catheter.

Regarding the clinical information at the time of echocardiography, it is important reporting mechanical ventilation strategies (see Fig. 2), and it is valuable to couple the study of RVF with information on LVSF and LVDF (or at least on estimation of LV filling pressure). Although not directly appraised, the value of pH and PaCO2 at the time of echocardiography exam may be valuable for their influence on pulmonary vascular resistances (and on catecholamines responsiveness).

LV diastolic function

The assessment of LVDF in the ICU population is a challenging task and it relies on the integration of several variables. However, there is evidence on the importance of LV diastolic dysfunction in critically ill septic patients [9, 10]. The most recent guidelines included the left atrium (LA) size and the estimation of PA systolic pressure—through the evaluation of the tricuspid regurgitant jet velocity—in their algorithm [11]. As such, the panel considered that the reporting of these two measurements is essential (Fig. 5). However, LA size may be particularly unreliable for acute changes of LVDF in critically ill patients undergoing mechanical ventilation with vasomotor and loading changes. Indeed, the LA is unlikely to dilate for acute worsening of LVDF [12]. If the authors choose to report PA systolic pressure, they should define how the pressure is obtained or calculated, e.g. obtained from PA catheter, calculated from the sum of tricuspid jet gradient and CVP or RAP (clearly stating how RAP is estimated).

The two other parameters derived from TDI (e′ and E/e′) are essential for the diagnosis of LVDF according to the new guidelines. The authors should clearly report if they investigated lateral, medial or average values of e′ velocity, since reference values are different according to the site of sampling [11]. Another essential parameter for grading of LVDF is the E/A ratio, and in case of significant tachycardia with merged flow, authors should clarify how they calculated it.

Pulmonary venous flow and E wave deceleration time are additional measurements, as not recognized by the current guidelines [11] although included in the previous ones [13].

The panel believes that when describing LVDF it is of utmost importance to provide information on the LVSF as patients with known LV systolic dysfunction have by definition impaired LVDF [10]. We believe it is very valuable to provide also data on RVF as since RV dilatation and eventually paradoxical septal motion can impair relaxation process and increase LV filling pressures.

In order to have a better clinical interpretation of echocardiographic measurements of LVDF, the panel felt that the authors should clarify the values of both systolic, mean and diastolic blood pressure. Although heart rate is an essential common item, the panel highlighted the value of providing this data as tachycardia may further impair LV relaxation during critical illness. The panel felt that it is essential to report not only current treatment with vasoactive drugs, but also longer term chronic cardiovascular drugs taken by the patients at the time of ICU admission. This knowledge may help to understand the burden of pre-existing LV diastolic dysfunction.

The panel judged that it is essential to indicate the technical details of measurement, criteria used for diagnosis and grading of LVDF, and to cite relevant references.

Fluid management

In studies assessing fluid responsiveness (FR), it is essential to clarify whether or not echocardiography was used as a reference standard. If so, it is essential to report the methodology of echocardiographic measurement to assess the variation of cardiac output (or its surrogate such as stroke volume or aortic velocity–time integral) after the intervention aimed at increasing venous return. The method used to increase venous return also remains essential in reporting, in particular the type of fluid challenge (nature of fluid, volume administered, duration of administration, safety limits) or the details of other manoeuvers such as passive leg raising. Authors should report technical details of the echocardiographic approach and the measurements performed to assess FR. When a new echocardiographic index is being developed, it is important to report the reference (gold) standard used for comparison.

The panel considers of paramount importance reporting cardiovascular and respiratory conditions for their profound influence on FR (Fig. 6). Although one must assume that FR is evaluated in patients in sinus rhythm, the cardiac rhythm should be clearly reported, and the timing of data acquisition within the cardiac cycle (i.e. end-expiration vs. end-inspiration) should be stated when evaluating stroke volume and velocity–time integral. RV size and presence of RV failure are also pivotal aspects since the ventricular interaction may lead to false-positive results regarding FR [14]. Respiratory parameters are essential as it is clear that FR is influenced by the mode of ventilation, tidal volume, respiratory rate and compliance.

Use of the recommendations

The PRICES recommendations are most useful when conducting prospective CCE research studies. Researchers are encouraged to collect and report “essential” and (possibly) “additional” items, according to the topic(s) investigated. On the other hand, PRICES recommendations should not refrain authors from publishing repositories of historical CCE information and these recommendations may be less useful in case of retrospective studies as some data might not be stored in the database. Therefore, we acknowledge potential challenges in the practical implementation of our PRICES guidelines. The goal of PRICES is not to exclude retrospective studies from the pathway of CCE research, but rather to provide researchers with some guidance in reporting, in the hope to limit biases in CCE research. In general, we believe that the PRICES recommendations are a useful guide as to what and which data to collect, if available. If unavailable, the researchers can simply report that some of the items were not available (for instance due to the retrospective study design).

Limitations

The recommendations presented in this document are the results of huge efforts of integrating information gathered from the systematic appraisal of the published literature subject to expert opinion [3]. We believe that this method is a significant strength of our approach but due to its novelty it may be subsequently refined. In this regard, we hope that it will be adopted and refined by other expert groups. Another limitation is that selected items were determined in advance and therefore some other items of interest may have not been considered.

However, we believe that the RI for these latter items would have not been rated as essential, but most likely as additional or optional. The authors fully acknowledge they did not entirely follow the approach suggested in the PRICES guidelines in their previous studies. Finally, this statement does not aim to guide researchers in the field of CCE on how they have to measure echocardiographic parameters but rather to give them a checklist of items that should be reported by prospective studies.

Conclusion

We report the PRICES recommendations, with an international consensus and a checklist for standardized reporting of CCE research studies. The intention is to provide guidance to researchers in the field of CCE with respect to methodology when developing research studies, and subsequently for a more standardized reporting of research findings.