Henry Blackburn on ‘The Minnesota Code for Electrocardiograms (ECG) in Population Studies’
During the 1950s, when most systematic population studies of heart disease began, it became important to have standard and quantitative means to compare cardiac disease rates. Diagnoses indicated on death certificates, or made by physicians with different training and language, might easily result in spurious differences. Even independent diagnoses by specially trained physicians were found highly variable. The need for comparability in clinical assessments led those in the Laboratory of Physiological Hygiene at Minnesota to consider use of the electrocardiogram as an objective measure for diagnosis and comparison. They early assembled and tested an ECG classification system and soon established collaboration with the London School of Hygiene and Tropical Medicine for evaluation of its criteria, eventually leading to systematic procedures and training of lay coders.
The electrocardiogram (ECG) indicates cardiac manifestations of greatest interest in epidemiology, that is, death and scars of heart muscle (infarction), inadequate blood supply (ischemia), increased muscle mass (hypertrophy), and disturbances of rhythm or conduction. The ECG seemed particularly promising to the Minnesota group as an objective graphic record amenable to standard procedures of collection, measurement, and classification. It also seemed ideal because it is acceptable, painless, simple, and inexpensive.
Ancel Keys assigned Henry Blackburn, then a young Research Associate who had just joined the staff in 1956, to develop an objective classification system for the ECG, along with the preparation of questionnaires, forms, and criteria to translate clinically relevant material into quantifiable form, all intended for the first round of cross-sectional surveys in the U.S. Railway Study and the Seven Countries Study beginning in 1957.
Several hurdles soon became evident. Differences among physicians’ blinded individual “impressionistic” reading of ECGs were large, even those by the same cardiographer reading at different times. Moreover, few objective criteria existed for abnormality or for specific cardiac findings; only gross pattern recognition and diagnostic labels were in use. The challenge of finding valid, repeatable, and standard criteria and measurement had first to be met before prevalent or endpoint cardiac “events” and their population rates could be reliably assessed and compared.
The first approach made was to compile existing electrocardiographic criteria and sort them into quantitative classes. Where these were inadequate, “reasonable” graded criteria were devised, validated, and empirically tested for sensitivity and specificity in well-defined populations that contained different proportions of normal subjects and patients with specific abnormalities. The criteria were then assembled, unambiguously described, rank-ordered by magnitude and amplitude and finally assigned code numbers. The resulting classes were quantitative, mutually exclusive, and directly relevant to common disease states. They were objectively described, however, as specific magnitudes of Q-QS waves, or negative T-waves, etc. rather than given clinically diagnostic labels of probable or possible “infarction, ischemia, or hypertrophy.” Continuous measurements were recorded of a few ECG amplitudes and intervals.
Finally, records from different living populations were classified and evaluated for “reasonableness” of the population distributions obtained, including sensitivity-specificity as applied to general, not in-patient populations. Test-retest reliability improved gradually among the resident physicians closely involved in development of the code: Pentti Rautaharju, Sven Punsar, Gunnar Blomqvist, and Henry Blackburn.
Between 1958 and 1959, the coded criteria were sent in an early version to other investigators involved in population studies. Ian Higgins at Cardiff, Fred Epstein in Michigan, and Geoffrey Rose in London became active collaborators in the testing. They began to use the criteria and provide suggestions for minor revisions, and soon became enthusiastic about an increasing ability to classify prevalence and incidence data quantitatively. The criteria and codes became known around this time as “The Minnesota Code.” [Note: Ian Higgins reports that it was very likely Aubrey Kagan of London, while he was on temporary assignment to WHO in Geneva in the late 1950s, who dubbed the system “The Minnesota Code.” This fortunate proprietary but impersonal name stuck.]
Actual records were extensively shared to assess variability in coding. Despite the use of unambiguous and quantitative criteria and mutual awareness of standard procedure, the coding variation among skilled observers remained large. Even clear definitions and objective criteria do not guarantee similar application by different observers.
Consequently, the collaborators gave greater attention to the conditions of electrocardiographic recording and measurement, and then to coding procedures and rules for dealing with ambiguities and differences. Measuring devices were developed with magnifying loops, particularly for assessing Q-wave duration and other intervals, which improved classification around the borderlines of criteria. Rules were devised to account for pattern variation among heart beats. The system was further shored up by quality-control procedures and standardization, with a system of duplicate, independent readings and adjudication of differences by a third party. The “final” Minnesota Code and procedure was published in Circulation in 1960, then modified slightly for the 1968 WHO method manual (Blackburn et al and Rose and Blackburn).
By the early 1960s, the burden of coding still lay mainly on Blackburn and a few visiting physicians in the Laboratory of Physiological Hygiene. Meanwhile, the volume increased as ECGs were sent to Minnesota for standard coding from many population studies. The high level of interest among the young physicians in training during the developmental phases of the coding system was superseded by boredom with what had become a tedious and routine (and unpaid) chore. At about the same time, independently, Rose in London and Blackburn in Minnesota hit upon coding by technicians. Rose approached the issue logically, unaccepting of any cardiological mystique, and began training of clerks. Blackburn’s resolution of the heavy reading load was serendipitous, arising during a special experience with ECG coding in his home. He recounts it here:
“In the summer of 1962, I was solicited by the National Health Examination Survey of the National Center for Health Statistics to classify electrocardiograms by the Minnesota Code on 6,000 individuals, a ‘true sample’ of the entire U.S. population. I agreed to do it as a summertime avocation. It offered the princely reward of 25 cents a record, but seemed otherwise ‘a worthwhile undertaking for science’—until I received the actual shipment of electrocardiograms. They came as unmounted strips stowed in tiny cardboard cubicles as tightly wrapped cylinders some six to nine feet long. These had to be teased from their cases, unrolled, held flat, read, measured, codified, tabulated, rerolled, and reinserted in the packing. The entire process took many times longer than reading, classification, and tabulation alone.
At the inducement of one cent per record, I was able to enlist the enthusiastic aid of two alert, very young, non-technical persons, relatives of mine who were close at hand that summer. While one, aged six, would extract and unroll the record, the other, aged seven, would hold it down until I coded it and then roll it back and refile it. Their summer wage came to 6,000 cents each, and mine, 6,000 x 23 cents. Their work reduced mine substantially, and as a side effect, nurtured a most pleasant relationship among us very near relatives.
After only a few days’ experience, from childlike curiosity, my young assistants could identify P, QRS, and T waves. After a little more time, they were careful to point out to me when a P wave or a T wave was ‘upside down’ or when a Q wave was ‘big and fat’ or an R wave was ‘too tall,’ and so on. After a few more days, by which time they spontaneously associated the code numbers I was writing down with tall, flat, inverted, or prolonged waves, it was clear that they could, at age six and seven, become excellent ECG coders!
Having read some 30,000 records a year for several years, this experience convinced me that it was time to call a halt. In fall of 1962, I began recruiting part-time university students and instructing them in the coding procedure. Many hundreds of student coder alumni have, over the four decades since, helped in this service to national and international population studies, while at the same time earning enough to help themselves through school.”
Blackburn adds: “It was not originally our intention to develop a coding system for export, rather to develop sound methodological procedures for assigning objective events related to cardiac disease for our extensive Minnesota-led surveys. But the Minnesota Code soon became widely used because it met a need in the burgeoning new field of cardiovascular disease epidemiology, particularly after it was published in Circulation and then included in the World Health Organization monograph series, Cardiovascular Survey Methods, published in 1968 (Rose and Blackburn).
Since then, the code has been expanded and updated to classify endpoints for clinical trials and to better characterize arrhythmias and conduction defects. A set of criteria was developed to estimate significant serial change in the ECG over time. Because the electrocardiogram has a different set of ‘errors’ than clinical assessments, and is independent of them, it complements clinical data from physicians who may inadvertently be biased by knowledge of drugs being administered in a trial, etc.
The original 1960 publication in Circulation was cited by the Cumulative Index Medicus in 1981 as one of the more extensively referenced articles in scientific literature. A manual of procedure for training and testing in the Minnesota Code was published in 1982” (Prineas, Blackburn, and Crow).
Electrocardiograms are now often recorded on magnetic tape, analog or digital, in field surveys and read by machine, including sophisticated new clinical diagnostic programs. Because of the need to compare these new with old records, software has been developed to identify Minnesota Code classes and they have proved quite reliable within the code’s original limitations of validity. The differing natures of man and machine pose both a problem and an advantage. The computer’s determination of wave onset and offset, and of baselines, though much more repeatable, is often systematically different from human-visual interpretation. Thus, the current strategy of the Minnesota group and of the Epicore Electrocardiograpic Center at Wake Forest University, is to use the computerized coding for things the computer does best (that is, measurement repeatability and averaging), and to complement its classification by things that the human does best (that is, judging baselines, resolving borderline and ambiguous situations, and coding complex arrhythmias). Combined human-machine adjudication achieves the more reliable and valid electrocardiographic classification.
For major new studies, completely automated measurement and classification are now available with NOVACODE, a system developed by Rautaharju, in large part from the extensive experience recorded in the Seven Countries Study, and with another developed by Jan Kors and his group in the Netherlands (Kors et al. 1996 and Rautaharju et al. 1990).
Few student coders at Minnesota go on to medical or health-related careers. However, the coding course allows students to become highly competent in a discipline that to this day remains something of a “mystique.” Coders become experts in detecting, differentiating, and classifying ECG findings. Rarely they may be in a position to misuse their electrocardiographic reading skills outside the research setting.
For example, we once had a student who tried to sell her skills as an ‘electrocardiographic coder’ to record rooms in Twin City hospitals. They, of course, had no idea what she was talking about and called us for fear that she was mentally unbalanced. The Minnesota Code is not standard in hospital practice. Rather, it is a system designed for rigorous population researches and clinical trials.
Later, we had another student who read her grandmother’s ECG as it was being made at bedside in the hospital, blurting out: “Oh, what a nice big 7-1 you have, Grandma. You have a complete left bundle branch block!” It took some hours of medical persuasion, plus sedation, to calm the elderly woman who thought she’d had a heart attack!” (Personal communication with author).
Blackburn, Henry, Ancel Keys, Ernest Simonson, Pentti Rautaharju, and Sven Punsar. 1960. The electrocardiogram in population studies. A classification system. Circulation 21: 1160-75.
Rose, Geoffrey and Henry Blackburn. 1968. Cardiovascular survey methods. WHO Monograph Series No. 56. Geneva: WHO Press.
Prineas, Ronald, Henry Blackburn, and Richard Crow. 1982. The Minnesota codemanual of electrocardiographic findings: Standards and procedures for measurement and classification. Boston: Wright-PSG.
Kors, J.A., G. van Herpen, J. Wu, Z. Zhang, R.J. Prineas, and J.H. van Bemmel. 1996. Validation of a new computer program for Minnesota coding. Journal of Electrocardiology 29 (Suppl): 83-8.
Rautaharju, P.M., P.J. Maclnnis, J.W. Warren, et al. 1990. Methodology of ECG interpretation in the Dalhousie program: NOVACODE ECG classification procedures for clinical trials and population health surveys. Methods of Information in Medicine 29: 362-74.
Henry Blackburn provides more background on early development of The Minnesota Code:
“Though I did all the compiling of the Minnesota Code, with the crucial aid of Punsar, Rautaharju, and Blomquist in testing it, a major source of data was the material that Ernst Simonson had acquired for his text on the normal and abnormal ECG (Simonson 1961). Our chief encouragers and collaborators in the early days were Fred Epstein of Michigan, Geoffrey Rose of London, and Ian Higgins of Cardiff. But the original impetus was from Ancel Keys, when he commissioned me for the larger enterprise of developing field methods for the Seven Countries Study. He took an active role in sending early drafts of the code around the world to get the reaction of outstanding authorities with whom he was on a first-name basis. Some of the correspondence of historical interest is found in a personal memoir, It Isn’t Always Fun. A few of these letters are reproduced here from central players:
Fred Epstein was much involved from the outset and discussed the criteria with others, including his Michigan colleagues Park Willis and Franklin Johnson, the latter having developed the chest (V) leads of the standard ECG, still used 50 years later.
Fred Epstein wrote to Ancel Keys in 1957:
“I had an impression from Dr. Doyle’s [Joseph Doyle of the Albany Heart Program] correspondence that diagnostic criteria of the type proposed are not practical or even desirable. I disagree with this point of view, in all friendliness, and have said so with no little emphasis in my letter to him. In general, I would like to say, and I think this is very important, that we need not quarrel as to what each of us means by definite or possible disease. All we have to do is to agree to list our findings, irrespective of their meaning, in a standardized way. For this purpose, I believe that the categories you propose are not open to any major criticism though one might even out some minor points.”
Thus, Fred gave ammunition to return to my plan not to label the ECG classes with diagnostic terms; such terms having been added during the short period when the code was “adopted” by Keys and Simonson for their presentation at the Princeton Conference on epidemiological methods. Characteristic of those times, despite having done the work, I was considered too “junior” to represent myself or this coding effort at such an important international conference of experts. [Never mind!]
Correspondence was ongoing with the more outspoken skeptic, Joe Doyle of the Albany Study, and this letter to him from Fred Epstein contains the gist of what we were seeking. It illustrates the problem of acceptance of defined criteria for population comparisons by those perhaps not yet disabused of their clinical infallibility by test-retest exposures. A few studies, including Framingham and Albany, have never departed from their reliance on expert opinions untested for simple repeatability. Their resistance to objectivity has not significantly impaired acceptance of their results or overall contributions.
July 15, 1957
I am surely in agreement with everything you say in your statement, and if I may say so, it was a pleasure to see it said so elegantly and well. However, can we go beyond this and be more specific without getting into controversy and rules and regulations to which others may not subscribe, and which may prove, like all uniformity, stifling and sterile?
I think that the task is not quite as complicated in practice as it seems in theory. For one thing, one is at an advantage in longitudinal studies, inasmuch as one does not have to commit oneself entirely on the significance of a given finding at the time it is taken, it being one of the very purposes of these studies to determine the meaning of, in your words, ‘borderline abnormalities.’ As long as we don’t call them normal or abnormal, I believe we are safe.
I believe failure to make some attempt at standardizing the criteria might cause major inconsistencies when comparing results of different studies. Clearly, when one places a cutting point high enough to exclude most of the false positives, one increases the number of false negatives. I think this is not too bad from our particular point of view, since the lesser evil is to miss some true positives, the greater evil to include too many false positives. Due to the insensitivity of the ECG, one misses many instances of more or less advanced disease anyway. My feeling is that ECGs should not be too precisely standardized, and reading between the lines, this seems to be your feeling, too. I gathered the impression in Boston that most people, whether they wanted it or not, resort largely to pattern reading. This certainly appears to apply to Framingham. I rather think that we should at least make an attempt to go a little bit further than this. Otherwise, our advice would merely consist in giving the green light to everybody to go his own merry way.
Frederick H. Epstein,
Research Associate and Lecturer in Epidemiology.”
Keys initiated correspondence with many cardiologists internationally about early versions of the Minnesota Code. Here is a reply from Paul Dudley White, dean of cardiologists:
264 Beacon Street, Boston.
February 11, 1958
Thanks very much for your letter of February 7. Yes, I suppose it is a good idea to have a uniform system of reporting electrocardiographic findings. I shall refer the list to Conger Williams for comment and then return it to you. I have edited an item or two myself, for example, about the physiological effect of exercise in producing ST depression and the possibility of horizontal heart position giving a fair amount of left axis deviation as shown in the limb leads. Also, I have said that one should exclude a normal, narrow split of the R in V2 in the diagnosis of right bundle branch block. One might also add an item on elevated ST segments and on bigger, wide T-waves don’t you think?
cc. Conger Williams
An October 15, 1958 letter to Ancel Keys and me from Ian Higgins, then working with Archie Cochrane in the Cardiff MRC read, in part:
“I think we all agree that precise quantitative criteria for ECG interpretation that would be generally accepted in epidemiological comparisons are urgently needed. Your criteria form an admirable starting point. It might be possible to reduce the number of criteria by excluding any that are poorly reproducible. I hope you won’t consider these [detailed enclosed] criticisms as too outspoken. We are as anxious as you to decide on generally acceptable criteria and must congratulate you on the progress you are making in this direction.”
[Ancel replied to Higgins letter with greater than usual warmth and diplomacy: “We are delighted at the prospect of developing close cooperation with you and your group and I am sure that this will be valuable to progress in the epidemiology of heart diseases as well as it being personally pleasant and interesting.”]
Grant, then author of a popular new book on vectorcardiography at the time, and soon to become director of the National Heart Institute, was quoted by then institute director James Watt in a letter to Keys:
“Dr. Grant is quite impressed with these criteria and feels that they are the best that he has yet seen. He does have some specific suggestions to follow which he feels may make the criteria more accurate.”
Howard Burchell, then Senior Consultant in cardiology at the Mayo Clinic, replied on February 17, 1958:
“I appreciate your sending me the proposed system of reporting electrocardiographic findings, and I believe that the outline is a very excellent one. The actual incidence of [confounding] conditions is so small that I doubt they would seriously interfere with the statistical appraisal of the electrocardiographic abnormalities as indicating coronary disease in a population.”
A collection exists of generally supportive letters from other cardiological leaders of the day, including Kossmann, Kimura, Astrup, Karvonen, Chapman, Aravanis, Aaron Schaeffer, Milnor, Bronte-Stewart, Briller, Chapman, Kannel, and A.J. Thomas. Ancel Keys’s thoroughgoing effort to get the input of these experts surely facilitated eventual acceptance of the Minnesota Code, perhaps in that no one was in a position to lambaste it once they were on record as for it!
Burchell, Howard. Howard Burchell to Ancel Keys, February 17, 1958. Letter. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.
Epstein, Frederick. Frederick Epstein to Ancel Keys, 1957. Letter. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.
Epstein, Frederick. Frederick Epstein to Joseph Doyle, July 15, 1957. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.
Keys, Ancel. Ancel Keys to Ian Higgins, October 15, 1958. Letter. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.
Simonson, Ernst. 1961. Differentiation between normal and abnormal in electrocardiography. St. Louis: Mosby.
Watt, James. James Watt to Ancel Keys, 1958. Letter. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.
White, Paul. Paul White to Ancel Keys, February 11, 1958. Letter. History of Cardiovascular Epidemiology Collection, University of Minnesota, School of Public Health.