Burdick 8300 and 8500 ECG Physicians Guide Brief Format Statements Rev B

ECG INTERPRETATION CRITERIA PHYSICIAN’S GUIDE

BRIEF FORMAT STATEMENTS BURDICK 8300/8500 70-00821-02 B

Information in this document is subject to change without notice. Names and data used in the examples are fictitious unless otherwise noted. CE Mark Declaration The CE marking of conformity indicates that the device having this symbol on its immediate label meets the applicable requirements of the European Medical Device Directive. Trademark Information Cardiac Science, the Shielded Heart logo, Quinton, Burdick, and HeartCentrix are trademarks or registered trademarks of Cardiac Science Corporation. All other product and company names are trademarks or registered trademarks of their respective companies. Copyright © 2011 Cardiac Science Corporation. All Rights Reserved.

Cardiac Science Corporation 3303 Monte Villa Parkway Bothell, WA 98021, USA 800.426.0337 (USA) 425.402.2000 [email protected] www.cardiacscience.com ii

70-00821-02 B

EC REP

MDSS GmbH Schiffgraben 41 D-30175 Hannover Germany Burdick 8300/8500

Contents

Contents Chapter 1: Introduction Historical considerations... 1-2 ECG wave recognition ... 1-3 Diagnostic approach ... 1-8 Intended use of program ... 1-12 Program specific information ... 1-13

Chapter 2: Preliminary Comments STATEMENTS ... 2-2 LEAD REVERSAL/DEXTROCARDIA ... 2-3 RESTRICTED ANALYSIS ... 2-5 MISCELLANEOUS PRELIMINARY STATEMENTS... 2-5 PEDIATRIC ECG ANALYSIS ... 2-5

Chapter 3: Heart Rate *** EXTREME TACHYCARDIA *** ... 3-2 TACHYCARDIA ... 3-2 *** EXTREME BRADYCARDIA ***... 3-3 BRADYCARDIA... 3-3

Chapter 4: Intervals PR INTERVAL ... 4-2 QT INTERVAL ... 4-3

Chapter 5: Conduction Defects INTRAVENTRICULAR CONDUCTION DEFECTS... 5-2 WOLFF-PARKINSON-WHITE PATTERN ... 5-5

Chapter 6: Hypertrophy LEFT VENTRICULAR HYPERTROPHY ... 6-2 RIGHT VENTRICULAR HYPERTROPHY ... 6-5 BIVENTRICULAR HYPERTROPHY ... 6-8

70-00821-02 B

Contents

iii

Contents

Chapter 7: Myocardial Infarction STEMI CRITERIA ... 7-2 Q WAVE CRITERIA... 7-3 INFERIOR INFARCTION ... 7-5 LATERAL INFARCTION... 7-6 ANTEROSEPTAL MYOCARDIAL INFARCTION ... 7-8 ANTERIOR MYOCARDIAL INFARCTION ... 7-11 SEPTAL INFARCTION... 7-14 POSTERIOR MYOCARDIAL INFARCTION ... 7-16 ANTEROLATERAL MYOCARDIAL INFARCTION ... 7-17 EXTENSIVE MYOCARDIAL INFARCTION ... 7-19

Chapter 8: ST Abnormalities CRITERIA ... 8-2

Chapter 9: ST-T Changes Acute ST-T Changes... 9-1 CRITERIA ... 9-2 STATEMENTS (REASONS) ... 9-3 STATEMENTS... 9-4

Chapter 10: Miscellaneous ATRIAL ABNORMALITIES... 10-2 QRS AXIS DEVIATION ... 10-3 LOW QRS VOLTAGES ... 10-5 TALL T WAVES... 10-6 SUMMARY CODES ... 10-7

Chapter 11: Rhythm Statements *** SIGNIFICANT ARRHYTHMIA *** ... 11-2 DOMINANT RHYTHM STATEMENTS ... 11-3 SUPPLEMENTARY RHYTHM STATEMENTS ... 11-5

Chapter 12: Program Accuracy Databases ... 12-2 Interpretation accuracy... 12-6 Measurement accuracy ... 12-21

Appendix A: References iv

Contents

70-00821-02 B

1

Introduction Contents ◆

Historical considerations

1-2

◆

ECG wave recognition

1-3

◆

Diagnostic approach

1-8

◆

Intended use of program

1-12

◆

Program specific information

1-13

Note: Computer assisted interpretation is a valuable tool when used properly. No automated analysis system is completely reliable, however, and interpretations should be reviewed by a qualified physician before treatment, or non-treatment, of any patient.

Welcome to the latest edition of the Physician's Guide that accompanies the ECG interpretive program. The aim of this Physician's Guide is to provide a list of the criteria currently used for ECG interpretation in Burdick/Quinton electrocardiographs. The ECG interpretation criteria are part of a program for ECG analysis that has evolved over many years and which was developed by the University of Glasgow Division of Cardiovascular and Medical Sciences, Section of Cardiology based in the Royal Infirmary, Glasgow, Scotland. Burdick/Quinton has worked closely with the team at the University of Glasgow to adapt and update this program for use by its customers. While some criteria are traditional, others have been developed through research studies and the need to quantify what, for many ECG abnormalities, has essentially been a subjective analysis of waveforms. Feedback from users has also helped to shape the latest release of the software. For the first time, this Physician's Guide provides an extensive outline of the approach to wave measurement and interpretation used by the University of Glasgow program. This follows from a requirement of the IEC 60601-2-51 specifications relating to "essential performance of recording and analyzing single channel and multichannel electrocardiographs". As in the past, the Physician's Guide also provides a detailed outline of criteria used by the program and provides data on program performance.

Burdick 8300/8500

70-00821-02 B

1-1

Introduction

Historical considerations Methods for the automated analysis of ECGs have been under continuous development in Glasgow since the late 1960's. A separate document reviewing the more historical aspects from early work on semi automated analysis of one cardiac cycle from three orthogonal leads and from the 12 lead ECG recorded in four groups of quasi orthogonal leads through to fully automated analysis of 10s of the 12 lead ECG with all leads recorded simultaneously is available [1]. That document also refers to various studies in which the software has been involved through the years. The remainder of this chapter deals with the methodology used in the currently available version of our software, which has become known as the University of Glasgow program specially for Burdick/Quinton devices.

1-2

70-00821-02 B

Burdick 8300/8500

ECG wave recognition

ECG wave recognition The methodology for ECG waveform measurement is described in general terms in an earlier publication from the Glasgow Laboratory [2]. 10 seconds of ECG data is input to the software for analysis and all leads require to have been acquired simultaneously.

Preprocessing Initially, a 50Hz or 60Hz notch filter is applied to remove any AC interference, if such a filter has not already been applied within the acquiring device. The first stage of the analysis is to compute any missing limb leads from the minimum of two leads that need to be provided. The ECG data is then filtered to minimize the effects of noise. The next step in the analysis is to calculate a form of spatial velocity combining the first difference of each lead. From this function, the approximate locations of all the QRS complexes are derived. Allowance has to be made for pacemaker stimuli, which are ideally detected by the front end equipment and passed to the program in the form of a list of "spike" locations. Given the QRS locations, it is then possible to check the quality of the recording for noise and baseline drift. If the drift is excessive, it is removed by using a cubic spline technique to obtain, for each lead affected, the baseline trend, which is then subtracted from the original data. If the noise is excessive, it is possible to remove a whole lead from the analysis or alternatively, 5 seconds of all leads are removed either from the first or second half of the recording.

QRS typing Thereafter, the various QRS complexes are typed according to their morphology. An iterative process is used. Effectively, the first complex in lead I is compared with the second using first differences of each cycle. The comparison takes the form of moving one beat over the other and when the difference is minimal, optimal alignment is present. This alignment point is used for averaging as discussed below. If the difference between beats is less than a threshold value, they are deemed to belong to the same class. The procedure is repeated with the third beat being compared with the second and so on. If a new morphology is detected, i.e. if the threshold is exceeded, a new class is established. The procedure continues with five leads being used in the typing process.

Burdick 8300/8500

70-00821-02 B

1-3

Introduction

Selection of required QRS class If more than one class of beat is present, then a decision has to be made as to which morphology will be used for the averaging procedure, i.e. the cycle to be interpreted has to be selected. A complex logic is used for this purpose. It has to allow for a single normally conducted beat in the midst of demand pacemaker beats for example. It also needs to take account of the QRS durations of different beat classes, RR intervals to exclude extrasystoles, and to a limited extent, the number of beats in each morphological class. The net effect is to choose one class of beats, of a similar morphology, that are regarded as being conducted in the normal sequence through the ventricles.

Averaging All beats in the selected class are averaged so that 12 such beats, one from each lead, are then available. The "average" beat can be computed in several ways. Common to this are the alignment points detected when wave typing was undertaken. They are used as reference points in the averaging process. The average beat can be a straight average of all corresponding aligned points, it can be a median calculated from the same points or it can be a weighted average - the so called modal beat introduced into the program in 1977 [3]. In this version, the program uses a special representative beat formation developed by Cardiac Science Corporation.

Wave measurement From the 12 average beats, a single combined function is formed and a provisional overall QRS onset and termination is determined by thresholding techniques. The provisional onset and termination are then used as starting points for a search for QRS onset and termination within each individual lead. Basically the approach conforms to the recommendations of the CSE working party [4] (of which one of the Glasgow team was a member), which were published in 1985.

Figure 1-1: Varying choice of baselines - reproduced from [4].

1-4

70-00821-02 B

Burdick 8300/8500

ECG wave recognition

Figure 1-2: Baseline at the level of QRS onset as used by the Glasgow program reproduced from [4].

In each individual lead, the QRS onset is taken as the baseline and hence Q, R, S, R' waves are measured with respect to the QRS onset as shown in the accompanying figures from the CSE paper (see Figure 1-1 through Figure 1-4). Isoelectric segments at the beginning of a QRS complex, i.e. a flat segment between the provisional overall onset and the onset of an individual lead are excluded from the first component (Q or R) of the QRS complex as recommended by the CSE group. Similar considerations apply at the end of the QRS complex (see Figure 1-3). A sorting algorithm is then applied to all 12 onsets to determine the global QRS onset as follows. The earliest onset is excluded and the next onset that also lies within 20 ms of the next again is then selected as the overall onset. This ensures that any true outliers are excluded. The reverse process is used to find the overall QRS termination.

Figure 1-3: Illustration of isoelectric segments I and K - reproduced from [4].

Burdick 8300/8500

70-00821-02 B

1-5

Introduction

QRS components Within the QRS complex, the amplitude and duration of the various Q, R, S, R' waves are then measured. In keeping with the CSE recommendations [4], the minimum wave acceptable has to have a duration >8 ms and an amplitude >20 μV. With respect to global QRS duration, the Glasgow program measures QRS duration from the global QRS onset to the global QRS termination. This means that an isoelectric segment within one particular QRS complex by definition will lead to a shorter QRS duration for that lead compared to the global QRS duration.

Figure 1-4: Definitions for QRS end / ST junction - reproduced from [4].

ST segment The ST segment has several measurements made. Figure 1-4 shows the J point as used in the diagnosis of ST elevation myocardial infarction. However, measurements are also made at equal intervals throughout the ST segment, e.g. 1/8 ST-T, 2/8 ST-T, etc.

1-6

70-00821-02 B

Burdick 8300/8500

ECG wave recognition

P and T waves A search for the P wave is made in the interval preceding the QRS complex. A P wave may not always be found in certain arrhythmias. P onset and termination are found using a method involving second differences but the same P onset and termination is used over all 12 leads in view of the difficulty in detecting low amplitude P waves in many leads. P wave amplitude is determined with respect to the same baseline as for Q, R, S amplitudes, namely the QRS onset. This was found to be more reliable than fitting a straight line between P onset and P termination even in cases where the P wave was superimposed on the T wave in the case of a tachycardia. T end is determined for each lead using a template method. The global T end is derived in a similar fashion to the global QRS offset. The other components of the ECG waveform, namely the ST and T wave amplitudes, are also measured with respect to QRS onset. Thus, the ST junction and the various ST amplitude measurements, such as ST 60 and ST 80 as well as the positive and negative components of the T wave, are all measured with respect to the QRS onset. The reason for this is that it is the most straightforward approach to measurement.

Interval measurements With respect to intervals, the global QT interval is measured from the global QRS onset to the global T end. On the other hand, because the P onset is taken as being simultaneous in all 12 leads, the global PR interval measurement is from the P onset to the global QRS onset.

Normal limits The above methods were used to determine the normal limits of QRS waveforms from an adult database of over 1500 normals, published in Comprehensive Electrocardiology, 1989[5] and a pediatric database derived from 1750 neonates, infants and children, published in part in 1989[6] and 1998[7] and which will be published in much more detail in the next edition of Comprehensive Electrocardiology. These normal limits are essentially an integral part of the diagnostic software.

Burdick 8300/8500

70-00821-02 B

1-7

Introduction

Diagnostic approach Rhythm analysis The QRS onsets and terminations used in wave typing are transferred to the rhythm program together with the measurement matrix of the 12-lead ECG. These data are used in determining the rhythm interpretation [8]. Three leads only are used for rhythm analysis. These are selected on the basis of the Pwave amplitudes determined by the wave-measurement program acting on the average beat. Leads II and V1 are always chosen and a third lead is selected from I, III, aVF and aVR, depending on the P-wave amplitude. In general, the above applies in the presence of an expected sinus rhythm. If flutter has been detected in Lead II, then Leads III and V1 are the other two leads which would automatically be used. If no significant P wave was found in the average beat, such as would occur very often in atrial fibrillation or other arrhythmias, such as complete heart block, the leads selected for analysis are II and V1, with two different P-wave morphologies being adopted for the latter. Because P waves have a different morphology in different leads, the template used for P wave searching varies depending on the lead under consideration. For example, if lead aVF has been selected and the P-wave amplitude is predominantly negative, then the template used for P-wave detection would be that of an inverted P-wave having first a negative, followed by a positive gradient as exemplified in the first difference of the data. P-wave searching is carried out from the end of an RR interval, i.e., from just before QRS onset in a reverse direction to the approximate end of the preceding T wave. If in any particular RR-interval P-waves are found to be absent, it is possible to alter critical values in the template and repeat the search. If a single P wave is found, then it would be retained. If multiple P waves are found, then they would be ignored being regarded as almost in the noise of the ECG. A variety of special subroutines has been developed through the years for different purposes. For example, in complete AV dissociation the P waves would be regularly spaced, but with no relationship whatsoever to the QRS complex. For this reason, a subroutine would check the regularity of any P wave detected and make allowance for the fact that some may have been missed on account of being submerged in the QRS or T wave. A PP-regularity index can then be calculated and a decision made on whether regular P waves, which are dissociated from the QRS complex, have occurred. The data from the average beats are also used to assess the likelihood of sinus rhythm being present where, of course, a definite P wave would be found in the average cycle in the vast majority of cases of sinus rhythm. The overall strategy of the approach is to detect sinus rhythm as early as possible in the logic by looking for the presence of regular rhythm with a single P wave in each RR interval and an essentially high value of PR regularity. Of course, the latter would be 1-8

70-00821-02 B

Burdick 8300/8500

Diagnostic approach

found in abnormalities such as 2-1 AV block and the presence of multiple P waves must be eliminated prior to the diagnosis of sinus rhythm. However, if pure sinus rhythm has been found, then an early exit from the rhythm analysis can be made. In all other cases, a more detailed analysis of rhythm commences. Abnormalities of the PR interval are assessed both on the basis of the median cycle and the PR interval as measured by the rhythm program. In cases of an extremely prolonged PR interval as in First Degree AV Block, only the rhythm program would accurately detect the lengthened PR interval. A significant amount of work was done on the use of neural networks to attempt to improve the accuracy of determining atrial fibrillation [9] but ultimately it was found that deterministic methods were equally acceptable. Differentiation of atrial fibrillation with rapid ventricular response from sinus tachycardia with frequent supra VES still remains a difficult problem for automated techniques. Relatively recently, newer methods for enhancement of reporting atrial flutter were reported by the Glasgow group [10]. While logic for detection of saw tooth waves has always been present, the more recent logic adopted a threshold crossing technique combined with regularity of intervals between peaks resulting in an improvement in the sensitivity of reporting atrial flutter from 27% to 79%, with a specificity exceeding 98% in both cases.

Burdick 8300/8500

70-00821-02 B

1-9

Introduction

Morphological interpretation The diagnostic component of the software is capable of using age, gender, race, clinical classification and drug therapy within its logic. Experience has shown however that many staff, particularly nursing staff, will simply not take the time to input the appropriate measures to the software, even the age and gender of a patient which it is known are fundamental to accurate interpretation.

Figure 1-5: Replacement of a step function threshold with an exponential function. With the step function, K points are scored when the ECG measurement exceeds the threshold 'b'. With the exponential function, the score varies continuously.

The basic approach to interpretation is through the use of rule based criteria but relatively recently this approach has been enhanced in several ways. First of all, smoothing techniques were introduced [11] to try to minimize repeat variation in interpretations by avoiding the use of strict thresholds between abnormal and normal. In short, instead of a step function separating normal from abnormal an exponential or even a linear function between the normal and abnormal threshold value can be used as illustrated (see Figure 1.5). This is associated with a scoring technique whereby it can be seen that a small change in voltage for example results in a small change in score. In the case of multiple parameters, more complex combination rules apply as discussed elsewhere [12]. Neural networks have also been introduced for detection of abnormal Q waves. However, it was found in practice that these perform best in combination with deterministic criteria [13]. Electrocardiography has not stood still in recent years and new terminology such as ST elevation myocardial infarction (STEMI) has been introduced. The software acknowledges the newer terminology and a significant amount of work has been done to adapt the output appropriately[14]. Another example of newer terminology is that of the Brugada pattern of which has to be taken into account.

1-10

70-00821-02 B

Burdick 8300/8500

Diagnostic approach

The software makes extensive use of the age and gender of patients in reaching an interpretation. Continuous limits of normality have been introduced particularly for children and younger males while different equations for normal limits are used for males and females especially in the younger adult age ranges. To a certain extent, the race of a patient is acknowledged through lower limits of normal voltage for Asian individuals, for example.

Burdick 8300/8500

70-00821-02 B

1-11

Introduction

Intended use of program Diagnostic application The Glasgow Program is intended to provide an interpretation of the resting 12 lead ECG in all situations, whether this be in a hospital or primary care setting. It is capable of diagnosing all commonly recognized ECG abnormalities such as myocardial infarction (MI), including acute MI, ventricular hypertrophy, abnormal ST-T changes and common abnormalities of rhythm. Conduction defects and other abnormalities such as prolonged QT interval are also reported. The software is not designed for interpretation of exercise electrocardiograms. The software has been widely used in clinical trials (e.g. the West of Scotland Coronary Prevention Study [15]) and hence has had wide exposure to recording of electrocardiograms in all commonly required situations.

Intended population The Glasgow Program is intended for use in adults and children of any age from birth upwards. The Program makes significant use of the patient's age and gender and indeed operates at the level of days in the case of neonates [6],[16]. It is believed to be the only program that is based on normal limits derived using the algorithm itself with this applying to criteria for subjects of all ages, including neonates. Indeed, it is known that other developers utilize the Glasgow normal limits.

Intended location The Glasgow Program is intended to be used in the hospital or in a general physicians office, or in out of hospital locations such as an ambulance. It is able to accept details of the patient's name, age, gender, and automatically invokes the appropriate criteria and routines such as special logic for acute cardiac ischemia where necessary. There cannot be any difference in ECG appearances of acute myocardial infarction depending on the location of ECG recording - it is only the prevalence of the abnormality that will vary.

Diagnostic accuracy The program is designed to be as accurate as possible with the emphasis being, if anything, towards a high specificity given that the criteria are based on the normal limits already described. Nonetheless, the program has high sensitivity for detecting all cardiac abnormalities as is evidenced by the results presented in the following section. In short, the program aims for the highest sensitivity at a high specificity although there is always a trade off between one and the other.

1-12

70-00821-02 B

Burdick 8300/8500

Program specific information

Program specific information Introduction This version of the software includes criteria for ST elevation myocardial infarction or STEMI. Of course, criteria for Q wave infarction remain. Several new statements dealing with ST-T changes in the presence of ventricular hypertrophy have been added in this version. As for the previous release, all statements have been redesigned so that reason statements are no longer necessary to supplement the diagnosis. In addition, most of the diagnostic statements have been further shortened compared to earlier versions. It is believed this will make the output more user friendly; the use of upper and lower case characters should also assist in the review process.

Pediatric aspects The Physician's Guide also contains pediatric criteria. Such criteria are automatically invoked when age is less than 16 to 18 years depending on the particular criterion under consideration. A unique feature of the Glasgow program is that continuous equations of upper limits of normal measurements are used. In general, such measurements will increase linearly from birth to adolescence. An example is QRS duration, which has an upper limit of 80 milliseconds at birth increasing to approximately 115 milliseconds at 18 years of age. Some measurements may reach their adult value at less than 18 years of age. The pediatric criteria can make use of lead V4R when it is available. While this enhances the accuracy of ECG interpretation in this age group, the program will still function when the conventional 12-lead recording positions are used in children (although the use of V4R to the exclusion of V3 is preferred). Full details of how to select the appropriate lead configuration for input to the electrocardiograph can be found in the relevant electrocardiograph operating instructions chapter.

Burdick 8300/8500

70-00821-02 B

1-13

Introduction

Age related information Similarly, the mechanism for specifying the age of a patient can also be found in the operating instructions. In the case of neonates and infants, the age will be calculated in days if the date of birth is input (for ages up to 364 days). If an age is input in years only, the criteria will be less efficiently used because the continuous equations employed allow the advantage of utilizing the age in days or months (age in months can be used for ages over 1 year). For example, for a patient who is 1 year and 11 months old, an age of 23 months should be entered as opposed to an age of 1 year. Clearly, there will then be significant differences in the upper limit of normal using continuous equations based on the age in days or months compared to using an age in years. A similar concept has been previously introduced for dealing with the upper limit of normal voltages for the diagnosis of ventricular hypertrophy in adults and children. Instead of discrete limits being used for particular deciles of age, continuous age dependent limits are now used.

Smoothing Finally, with respect to the new methodology, a major effort has been made to minimize the effect of diagnostic thresholds in ECG criteria. There has to be a border between normal and abnormal but the newer approach tries to smooth such boundaries by, for example, awarding only a small increment to a score if a measurement exceeds a threshold by only a small amount.

Presentation of criteria A principal aim in preparing this manual was that the criteria should, wherever possible, be presented in a relatively straightforward form. At the same time, it was intended that the text should convey the unique flavor of the approach used for ECG analysis within Burdick/Quinton brand devices. For this reason, a compromise has been adopted where, for some criteria, a generalized statement has been made rather than a precise quantification of numerical data being listed. Even so, the list of criteria is somewhat detailed but it should be appreciated that computers do require a certain amount of precision! Detailed criteria for arrhythmias are not listed although the various statements that can be produced by the program are presented. The layout of the criteria should be self-explanatory. In general terms, the wave amplitudes have positive or negative amplitudes in the conventional sense, e.g. the S wave is regarded as having a negative amplitude. Similarly, a criterion which requires that T- < -0.1 mV means that the negative T wave amplitude should be in excess of -0.1 millivolts, e.g. this would be true if T- = -0.2 mV (see Figure 1-6 on page 1-16). Occasionally, the absolute value of a negative wave or ratio is denoted by | |.

1-14

70-00821-02 B

Burdick 8300/8500

Program specific information

Use of clinical data A unique feature of the program is the ability to make use of age, gender, drug therapy and clinical classification of the patient. The full list of clinical classifications that may influence the interpretation is as follows: ◆

Normal

◆

Myocardial Infarction

◆

Myocardial Ischemia

◆

Hypertension

◆

Congenital Heart Disease

◆

Rheumatic (Valvular) Heart Disease

◆

Pericarditis

◆

Respiratory Disease

◆

Implanted Pacemaker

◆

Endocrine Disease

◆

Pulmonary Embolism

◆

Post-operative Cardiac Surgery

◆

Cardiomyopathy

◆

Other

◆

Unknown

The list of drug therapies accepted is as follows: ◆

Digitalis

◆

Diuretic

◆

Beta Blocker

◆

Quinidine

◆

Procainamide

◆

Amiodarone

◆

Disopyramide

◆

Lidocaine

◆

Other antiarrhythmics

◆

Psychotropic drugs

◆

Steroids

◆

Calcium Blockers

◆

NitratesAce Inhibitors

◆

Alpha Blockers

Burdick 8300/8500

70-00821-02 B

1-15