Ultrasonography in the Emergency Department

Protocol RCT 1.4 Ultrasonography in the Emergency Department A Randomised Controlled Multicenter Study RCT 1.4.1

Background Patients arriving to the Emergency Department with acute symptoms of respiratory problems are a diagnostic challenge. Their symptoms can be very similar and at times life threatening but require different specific treatments. The assessment of a fast and correct diagnosis is therefore of high importance in these patients where a correct and instant treatment can be lifesaving.

In 2014 Lancet Respiratory Medicine published an article written by Laursen et al. They presented at Danish study that showed that by using a focused ultrasonographic examination as a complementary tool to the regular diagnostic assessment of patients with respiratory symptoms they increased not only the percentage of patients receiving a correct diagnosis but also the percentage of patients receiving correct treatment within 4 hours after admittance to the Emergency Department.

The study was criticized for being performed by a single physician, who is a very experienced ultrasonographer at a single center and therefore not transferrable to the Danish Emergency Departments and their physicians in general.

Consequently this study is an expansion of the study of Laursen et al. accommodating some of the points of criticism. It will be executed as a multicenter study in several Danish Emergency Departments. The ultrasonographic examination will be performed and interpreted by the medical physicians who already work at the emergency departments.

Purpose To investigate if the percentage of correctly diagnosed patients 4 hours after admittance to the Emergency Department (ED) is increased when the patients are examined and diagnosed using standard diagnostics and focused ultrasonography of the lungs and heart, compared to using X-ray alone in a multicenter study.

Patients included Patients arriving to the internal medicine ED with difficulties of respiration will be screened for participation.

Hypothesis That there is no difference in diagnostic accuracy 4 hours after admittance to the ED when comparing the usual diagnostic assessment to usual diagnostic assessment + ultrasonographic examination (US) of the lungs and heart.

Trial design A prospective randomised controlled clinical multicenter study with parallel group design. Allocation ratio 1:1. A superiority study.

Gold standard for the primary endpoint The gold standard for a correct diagnosis at 4 hours after the patients arrival is defined as the final diagnosis achieved by medical journal audit made once the patient is discharged from hospital. This is done by two auditors. In case of disagreement a third auditor will repeat the medical journal audit.

Methods Participating centers EDs at the following hospitals will be participating: OUH, OUH-Svendborg, , Esbjerg, Åbenrå, Viborg, Ålborg, Køge, Holbæk, Hvidovre.

The project has been approved by the Ethical Committee of the Region of Southern Denmark and the heads of the EDs from each of the participating hospitals.

Investigators The investigators are medical physicians who are either specialists or physicians in specialist training who regularly attend to patients in the ED and who are familiar with US as a supplemental tool in the assessment of a medical diagnosis. The project manager is responsible for educating the investigators in the projects US Protocol as well as supervision will take place if necessary.

Project Ultrasound protocol

This protocol consists of an US of the lungs and the heart:

The USexamination of the is performed as follows: The anterior and lateral part of thorax is divided into a superior and inferior quadrant. Each quadrant represent a zone in which the probe shall be placed centrally and create a transverse picture of the costae and pleura. A sequence of at minimum 6 seconds including at least one inspiration and one expiration shall be recorded in each and every zone.

The US examination of the heart is inspired by the FATE protocol from which we use a 4-chamber picture of the heart achieved either from a subxiphoid or an apical window. A sequence of minimum 6 seconds of the heart shall be recorded.

All recordings from the US will be saved as raw data for later evaluation by specialists in US, blinded with respect to the investigators evaluation of the US as well as to the presumptive and final diagnoses of the patients.

Randomization Conducted as block randomization using a random-number generator. The investigators will have external access to the database, from which they continuously will be assigned a randomization number.

Blinding The physician is blinded to the randomization result until the primary tentative diagnosis is registered. The results of the US will only be unblinded for the treating physician in the intervention group. Unblinded US results are registered by the investigator on a registration chart and archived in the ED labeled with the patients ID-number.

In the electronic patient journal system we do not register randomization number nor the US results. Details registered in the electronic patient journal system will be the name and registration number of the project, the acceptance from the patient, name of the local investigator performing the US, and the localization of the registration chart in the ED containing the results of the US. We hereby create the possibility of a blinded audit using the electronic patient journal and we make it possible to obtain the results of the US in case the results are needed later in the diagnostic process and the investigator is not present.

Data collection and inclusion of patients The investigator will be screening all patients admitted to the ED according to inclusion and exclusion criteria. He will obtain informed consent both verbal and in writing from habile patients who meet the inclusion criteria. The investigator fills out the registration charts in the assigned database.

The attending physician performs a primary medical assessment of the patient and fills out the registration chart of Presumptive diagnoses at primary assessment. The investigator unblinds the randomisation number and performs the US examination according to the projects US protocol. The investigator informs the physician in charge of the primary medical assessment about the ultrasound findings on patients in the intervention group. He then fills out the registration chart Ultrasound findings. The investigator fills out the registration chart Ultrasound findings on patients in the intervention group and archives it in a map located in the ED.

The patient continues treatment and further investigation according to the department guidelines.

Once the patient is discharged from hospital the following will take place:

1. Blinded review of the US.
2. Registration of the correct diagnosis by blinded audit of the electronic medical journal.
3. Registration of the basic characteristics and of data for secondary endpoints.

Risks, side effects, and disadvantages There is no risk of radiation and no known side effects in relation to the US.

Enforceability of the project Main supervisor: Professor G Baatrup. Research Responsible in the Department of Surgery, Odense University Hospital (OUH)-Svendborg.

Co-supervisor: MD, Ph.D, MiAH, Clinical Associate Professor. L. S. Teglbjærg. Responsible of investigation in Emergency Care in the ED/Department of Internal Medicine OUH-Svendborg.

Co-supervisor: Ph.D., MD, C. B. Laursen. Department of Internal Medicine (Pulmonary Diseases) Dep. J, OUH. Certified in focused ultrasonographic examination. Ph.D. thesis on focused ultrasonographic examination at the ED at OUH. Chairman of the Ultrasound Committee the Danish Association of Emergency Medicine

Collaborator: Clinical professor in Emergency Medicine, Consultant, MD, A. T. Lassen. Head of Research Unit of Acute Medicine at the ED at OUH and Head of the Network of Researchers in Emergency Care in Southern Denmark.

Sample-size Is based upon an estimate of the percentage of patients expected to receive a correct tentative diagnosis within 4 hours after the arrival to the ED. This estimate is set to 65% based on data from Laursen et al. A clinically relevant enhancement of the tentative diagnosis by using US is set to a minimum of 15%. To detect a 15% increase in the number of correct tentative diagnoses, from 65% in the control group to 80% in the intervention group, with an 80% chance for detection and a level of significance of 5% 272 included patients are needed. The estimated drop out is 6% and increases the number of inclusion to 288 patients.

Data analysis The gold standard for diagnostic accuracy is constituted of a medical journal audit including the results from both US and chest X-ray examination in both groups of patients.

DATA ANALYSIS Descriptive data Descriptive statistics for both groups will be given; health characteristics; patient symptoms; measured variables upon arrival to the ED as well as answers to the primary and secondary endpoints. Categorical data will be summarized using number and proportion of patients, while continuous data will be presented using the number of patients (n), mean, SD, median, minimum and maximum.

Primary end point The c2 test, alternatively the Fischer exact test will be used to establish if there is a difference in the distribution of the total number of patients with a correct/incorrect '4 h' presumptive diagnosis in the control group and in the intervention group.

Secondary end points To compare the intervention group with the control group we will use the following tests: for the comparison of means: the Student t test; for the comparison of medians; the ManneWhitney test and for the comparison of proportions; the c2 or the Fisher exact test. All tests will be performed with a two-sided significance level of 5%. The gold standard will include the UL examinations evaluated by specialists in both groups. Using the audit diagnosis as the gold standard, for both groups, we will assess the diagnostic performance of the primary evaluation, the '4-h' presumptive diagnosis and the UL examinations by calculating sensitivity, specificity, positive predictive values, negative predictive values and diagnostic accuracy and their 95% CI. We will analyze data using the intention-to-treat principle. Data analysis will be conducted using STATA (StataCorp).

Data entry and security Measured data are entered into the database made for the project in RedCap OPEN. All data are stored and managed according to the laws and regulations as stated by the Danish Data Protection Agency.

Time frame Based on a pilot study the estimated period of inclusion will be 1- 5 weeks based upon a duty frequency varying in between 1 to 5 8 hours shifts a week and an estimated inclusion rate of 20 patients per investigator. The period of inclusion is set to 16 weeks.

Ethics The study will be conducted in accordance with the Helsinki II Declaration and as approved by the Regional Ethics Committee and the Data Inspectorate. In case of UL findings compatible with an acute life-threatening condition the attending physician will be informed and UL findings unblended. This will be registered in the database.

Procedure for collecting informed consent The project will comply with the guidelines issued by The National Ethical Scientific Committee, law nr. 593 of 14th of June 2011 concerning processing ethical research of research projects in the field of health science.

The investigation of the diagnostic features of the UL examination in the primary care demands short time from delivery of written and verbal information to collection of informed consent and the examination itself. Allowance for reduced consideration time for the patient has been approved by the Regional Ethics Committee of Southern Denmark.