Population Versus Clinical View of Case Fatality From Acute Coronary Heart Disease
Results From the WHO MONICA Project 1985–1990
Background The clinical view of case fatality (CF) from acute myocardial infarction (AMI) in those reaching the hospital alive is different from the population view. Registration of both hospitalized AMI cases and out-of-hospital coronary heart disease (CHD) deaths in the WHO MONICA Project allows both views to be reconciled. The WHO MONICA Project provides the largest data set worldwide to explore the relationship between CHD CF and age, sex, coronary event rate, and first versus recurrent event.
Methods and Results All 79 669 events of definite AMI or possible coronary death, occurring from 1985 to 90 among 5 725 762 people, 35 to 64 years of age, in 29 MONICA populations are the basis for CF calculations. Age-adjusted CF (percentage of CHD events that were fatal) was calculated across populations, stratified for different time periods, and related to age, sex, and CHD event rate. Median 28-day population CF was 49% (range, 35% to 60%) in men and 51% (range, 34% to 70%) in women and was particularly higher in women than men in populations in which CHD event rates were low. Median 28-day CF for hospitalized events was much lower: in men 22% (range, 15% to 36%) and in women 27% (range, 19% to 46%). Among hospitalized events CF was twice as high for recurrent as for first events.
Conclusions Overall 28-day CF is halved for hospitalized events compared with all events and again nearly halved for hospitalized 24-hour survivors. Because approximately two thirds of 28-day CHD deaths in men and women occurred before reaching the hospital, opportunities for reducing CF through improved care in the acute event are limited. Major emphasis should be on primary and secondary prevention.
Coronary heart disease (CHD) mortality rates vary markedly between countries,1,2 reflecting both the variation in incidence and case fatality (CF). The WHO MONICA Project (Multinational MONItoring of Trends and Determinants in CArdiovascular Disease)3 has collected data continuously on all suspected cases of acute myocardial infarction (AMI) or coronary death over a 10-year period using population-based registers of acute coronary events. Many studies have considered CF only for hospitalized AMI cases. An important advantage of the MONICA Project’s population-based registers is that the population view of CF can be obtained as all acute episodes of CHD, including those resulting in death out of the hospital, can be identified, thus considerably increasing the estimated percentage of CHD events that are fatal. The MONICA database also provides an opportunity to study the relationship between CF and coronary event rate for 29 populations. The large numbers of events in widely differing populations allow the examination of CF by different time periods (out of hospital, in hospital) and for first and recurrent event and permit an integration of the clinical and population view of CF.4–8 The long-debated potential differences in CF between men and women can also be addressed by analysis of CF by sex and age.9–11
The registration methods used have been described previously.5,6 All suspected acute coronary events, including out-of-hospital coronary deaths, in people 35 to 64 years of age, have been registered in a standardized way. Nonfatal events have been classified into one of the MONICA diagnostic categories on the basis of coding of reported symptoms, enzyme levels, and ECG findings.5,6 For fatal events, necropsy findings, if any, and history of CHD have also been used to determine the MONICA diagnosis. Detailed criteria and quality assurance procedures for event registration have been published.5 For this cross-sectional study, all coronary events in either of the following categories have been included5,6: (1) nonfatal definite AMI (NF1) according to MONICA diagnostic criteria and (2) deaths satisfying the MONICA definitions for fatal definite AMI (F1), possible coronary death (F2), or unclassifiable (F9) (called insufficient data in the manual6).
Thus all 79 669 events of definite AMI (NF1) or coronary death (F1,2,9) occurring from 1985 to 1990 among 5 725 762 people, 35 to 64 years of age, are the basis for CF calculations. Data from 29 populations that met data quality criteria by October 1995 were used (Table 1⇓). (Nine of 38 populations in the MONICA Project were excluded from this report: Registration of fatal events was incomplete in three populations; in one population, data were inadequate to estimate CF; one had not registered events for enough years; the quality of survival time data were insufficient in two; and two had errors in coding events as hospitalized. Data are for the years 1985 to 1990 except for Bremen, Germany (1985–1989), Friuli, Italy (1985–1989), and Northern Sweden (1986–1990).
Definitions of Case Fatality
Events have been classified into four groups according to survival: group 1 (G1), prehospital deaths, ie, persons who died before hospitalization; group 2 (G2), patients who were hospitalized and died within the first 24 hours after onset of symptoms; group 3 (G3), patients who were hospitalized and survived the first 24 hours but died within 28 days from onset of symptoms; and group 4 (G4), patients who were hospitalized and survived 28 days after onset of symptoms.
From these groups, five definitions of CF are derived (Table 2⇓). These vary both in terms of the end of the period of observation, 24 hours or 28 days, and the beginning, considering all events, only hospitalized events, or only hospitalized events where the person survived 24 hours after onset of symptoms. CF was defined as the number of fatal events as a percentage of the total number of fatal plus nonfatal events.
Coronary events are classified as first or recurrent AMI using information about history of previous AMI. Although this information is available for most hospitalized events, it is frequently lacking for prehospital deaths. Therefore, CF for first versus recurrent event is analyzed only in hospitalized events and those with unknown history of previous AMI were omitted from the calculations.
Direct age standardization of the event rates and the mortality rates was done according to the procedures of the WHO MONICA Project,2,6 using the truncated Segi world standard population weights 6, 6, 6, 5, 4, and 4 for the 5-year age groups from 35 to 39, to 60 to 64 years, respectively.12
Age-standardized CF was calculated with direct standardization for the 10-year age groups 35 to 44, 45 to 54, and 55 to 64 years, with weights of 1, 3, and 7, on the basis of the age distribution of events for the WHO MONICA Project as a whole. The age-standardized CF for a population is given by 100(∑wi[di/ei]), where i=1, 2, 3 indicates the age groups, wi denotes the weight, ei is the number of events, fatal and nonfatal, occurring in age group i for that population, and di is the number of fatal events. Approximate 95% confidence intervals are given by 100∑wi(di/ei)±1.96 SE, where SE2=1002(∑wi2di[ei−di]/ei3).
The relationships between CF and CHD event rates or age were explored by means of graphs and simple linear regression, separately for men and women. Sex differences in CF were analyzed by use of women-to-men CF ratios as the dependent variable (to focus on relative as opposed to absolute differences) and age and event rates as possible explanatory variables. The analysis of the variation of age-specific CF was mainly graphical.
Table 1⇑ describes the 29 populations in terms of mean population sizes, numbers of fatal and nonfatal CHD events, CHD event rates (fatal plus nonfatal), and CHD mortality rates for the age group 35 to 64 years in 1985 to 1990. The age-standardized CHD mortality rates (fatal events per 100 000 persons per year) based on the MONICA registers had an enormous range, from 46 in Beijing (China) to 395 in North Karelia (Finland) for men and from 15 in Catalonia (Spain) to 127 in Glasgow (UK) for women. The CHD event rates ranged from 79 to 818 for men and from 33 to 267 for women.
Fig 1⇓ shows sex-specific age-adjusted percentages of all 28-day CHD deaths that occurred before hospitalization; in the hospital within 24 hours after onset of symptoms; and between 24 hours and 28 days after onset of symptoms, among hospitalized events. For women, 42% to 75% (median 64%) of all CHD deaths within 28 days occurred before reaching the hospital. This was somewhat lower than for men, for whom the range was 58% to 80% (median, 70%). Moreover, this percentage was lower for women than for men in 22 of the 29 populations.
Tables 3⇓ and 4⇓ show age-adjusted CF by various stages of the acute event. Among hospitalized events, 28-day CF was 15% to 36% (median, 22%) for men and 19% to 46% (median, 27%) for women. However, when prehospital events were included, overall 28-day CF was substantially higher: 35% to 60% (median, 49%) for men and 34% to 70% (median, 51%) for women. Hospitalized patients who survived 24 hours after onset of symptoms had much lower 28-day CF: 8% to 21% (median, 14%) for men and 11% to 30% (median, 16%) for women. Age-adjusted 24-hour CF was 29% to 51% (median, 41%) for men and 22% to 60% (median, 42%) for women.
Age-adjusted 28-day CF for hospitalized events was higher for women than men in all 29 populations, by as much as an absolute 12%. However, prehospital CF was higher for women in only 13 populations. The combination of these different periods resulted in age-adjusted overall 28-day CF being higher for women in 21 out of 29 populations.
Fig 2⇓ shows age-adjusted CF by several definitions plotted against age-adjusted CHD event rate over the 29 populations by sex. For 28-day CF, there is no relationship for men and a negative relationship for women. For example, by linear regression it was estimated that for populations with an age-adjusted coronary event rate 50 per 100 000 higher, the age-adjusted 28-day CF for women would be an absolute 4.3% lower (P=.03). The relationship remained negative and statistically significant for women for all other definitions of CF except in 28-day CF for hospitalized 24-hour survivors. For men there was no association with event rate for any definition of CF. The graphs show that the strong relationship observed for women between 28-day CF and CHD event rate was mainly due to an association between 24-hour CF and CHD event rate.
The unadjusted and age-adjusted ratios of CF for women to CF for men are shown in Table 5⇓ for 28-day CF and in Table 6⇓ for 28-day CF for hospitalized events. The ratios were all larger than unity for hospitalized events (median, 1.32); they were always smaller for overall 28-day CF (median, 1.09) than for 28-day CF for hospitalized events. These ratios were generally reduced by age adjustment, but even then the sex ratios in 28-day CF for hospitalized events remained >1.20 and statistically significant for over half the populations (median, 1.24).
There were statistically significant inverse associations between the women-to-men ratios of age-adjusted CF and the mean (men and women combined) age-adjusted CHD event rates for all definitions of CF (Fig 3⇓).
Fig 4⇓ depicts the mean (all populations) 28-day CF with 95% CI, by 5-year age groups, for men and women. In men, the mean CF decreased from the 35-to 39-year age group to the next by an absolute 3% (P=.26 by t test). There was little change to the next 5-year age group, but from the 45- to 49-year to the 60- to 64-year age group, there was a 5% to 7% absolute increase in CF with each 5-year age group (each with P<.02). Although the shape of the curve for women is similar to that of men, the increase between age groups 40 to 44 and 50 to 54 was small. Between age groups 35 to 39 and 40 to 44, the mean decrease was an absolute 5% (P=.25) and between age groups 55 to 59 and 60 to 64, the mean increase was 6% (P=.03). Further, CF was higher for women than men in 22 of the 29 populations at age 40 to 49 years, with a mean CF of 39% for men and 49% for women. However, in the 55- to 64-year age group the mean 28-day CF was nearly identical for men and women, being 54% and 55%, respectively.
Table 7⇓ shows CF for hospitalized events for known first AMI and for known recurrent AMI. The percentage of events for which AMI history was unknown was 0% to 33% (median, 2%) for men and 0% to 38% (median, 2%) for women. Events for which the determination of first versus recurrent could not be made were excluded. CF was about twice as high for recurrent events as for first events. The median of ratios of recurrent event to first event for 28-day CF for hospitalized cases was 2.1 for men and 1.7 for women. (These are medians of the individual population ratios [not shown], not ratios of the medians.) The medians of these ratios for hospitalized cases who survived 24 hours were 2.2 and 1.9 for men and women, respectively. Women-to-men ratios for 28-day CF were higher for first than recurrent events: 1.34 and 1.17, respectively, for all hospitalized cases, and 1.44 and 1.26, respectively, for hospitalized 24-hour survivors.
The value of international comparative studies such as this depends on the extent to which the data are truly comparable. Collection of carefully defined data items, standardized diagnostic criteria, and regular quality assessment in the WHO MONICA Project are strengths of the present study. Once a coronary event is registered and diagnostic information collected, the classification of nonfatal events is likely to be the same in all centers, provided that the proportion of missing data for the diagnostic variables (symptoms, ECGs, and cardiac enzymes) is small. Similarly, with complete information for fatal events on history of CHD, symptoms, and necropsy findings, classification of the events is likely to be similar at all centers. However, the percentage of fatal events with unclassifiable data was ≈25% higher for women than men and was inversely associated with CHD event rate.5
The assessment of sex differences in CF has received much attention9 but is greatly affected by the choice of definitions and by whether the study is population based or hospital based. Vaccarino et al9 reviewed 27 studies of sex differences in 1-month CF for hospitalized AMI events (with sufficient enzyme or ECG data to ascertain AMI). They found higher CF levels in women than in men in all but one study. Since mean age is greater for women CHD patients than for men, and CF is associated with age, it would be expected that age adjustment would reduce these differences. Indeed, in the 11 studies that presented both unadjusted and age-adjusted results, the age adjustment reduced the women-to-men ratio of CF to <1.2 and statistical nonsignificance in all but two studies. Although the age adjustment did generally reduce the sex differences in the 29 MONICA populations, the women-to-men ratio of 28-day CF for hospitalized events remained >1.2 and statistically significant in over half of the populations. In the MONICA populations, the women-to-men ratios of overall 28-day CF (including CHD deaths occurring before hospitalization) are closer to those reported by Vaccarino et al, namely <1.2 and statistically significant for only 5 of the 29 populations.
For both men and women, CF was higher in the 35- to 39-year age group than in the 40- to 44-year age group, reflecting either more severe disease or slower recognition and treatment of younger CHD patients. The difference between men and women in overall 28-day CF almost vanished by age 55 to 64 years, indicating the need for qualifying the statement that women have higher CF than men.
The WHO MONICA Project has also shown that in men CF was not related to the absolute magnitude of the CHD event rates in a population, but in women CF decreased with increasing CHD event rates. An interpretation of these differences is that some nonfatal AMI events are missed in populations with very low CHD event rates for women.5 A recent paper from the Glasgow MONICA center,11 which has high event rates and fairly complete data on previous CHD for out-of-hospital deaths, did not find significant CF differences between men and women at any age for overall 28-day CF, supporting the suggestion that higher CF for women in other populations could be due to failure to recognize nonfatal AMI cases. The Scottish paper however, as here, showed a significantly lower percentage of 28-day deaths in women than in men occurring before hospitalization and a significant excess of deaths afterward. More higher-risk women than men are surviving long enough to reach the hospital alive. This is why CF for hospitalized cases is higher in women than in men.11
The CF differences shown here between the several definitions serve to highlight the difference between population-based studies, which include prehospital deaths, and hospital-based studies, ie, the population versus clinical view of CF. Coronary heart disease CF of 40% to 50% is typically reported from community studies of AMI.13,14 In contrast, hospital-based studies that include deaths in emergency rooms and typically use CF definitions similar to ours for hospitalized events obtain CF of about half the size of our overall 28-day CF.15–18 Coronary heart disease CF in cases who survive long enough to have data available on enzymes and ECG, ie, who survive several hours after reaching the hospital, is lower and may be closer to our CF for hospitalized 24-hour survivors.19
The variation across populations was least for overall 28-day CF. This may be due to the fact that other CF definitions are affected by how quickly people get to the hospital or receive thrombolytic therapy and by difficulties in classifying cases who die soon after arrival at the hospital. This also suggests that prehospital CF plus 24-hour CF for hospitalized patients is more valid for cross-population comparisons of early CF than prehospital CF. Variation in CF across populations was least for overall 28-day CF, which suggests that differences between populations in emergency care and transport to the hospital affect time of death rather than overall 28-day CF.
A great advantage of the WHO MONICA Project is its ability to provide both a population and a clinical perspective in the study of CF. In men, a rule of thumb might be that the overall 28-day CF is halved for hospitalized events and again nearly halved for hospitalized 24-hour survivors. This holds even among populations with widely differing 28-day CF. The respective percentages for women are slightly higher but are similar. In light of these population and hospital-based CF data, studies reporting very high survival figures for AMI patients should be put into perspective. They are likely to have investigated acute events rather late in their natural history, eg, when AMI patients have survived at least 24 hours in a coronary care unit.
Of all CHD patients who die within 28 days after onset of symptoms, about two thirds die before reaching hospital, therefore opportunities for reducing CF through improved care during the acute event are limited, emphasizing that the CHD epidemic can best be conquered by strengthening primary and secondary prevention for CHD.
Sites and Key Personnel of Contributing MONICA Centers
University of Western Australia, Nedlands
M.S.T. Hobbs,4 K. Jamrozik,4 R.W. Parsons, R. Broadhurst, C. Spencer, P.L. Thompson
University of Newcastle
A. Dobson,1 S. Leeder,2 H. Alexander, R. Heller, R. Gibberd, P. Steele
Dalhousie University, Halifax, Nova Scotia
R.D. Gregor,1 B.R. Mackenzie, P.M. Rautaharju, H.K. Wolf1
Beijing Heart, Lung, and Blood Vessel Research Institute, Beijing
W. Zhaosu,1 W. Yingkai,2 Y. Chonghua, H. Zhaoguang
Copenhagen University Hospital, Glostrup
M. Schroll,1 M. Kirchhoff, B. Hansen, S.Q. Lund, H. Schnack, M. Madsen
National Public Health Institute, Helsinki
J. Tuomilehto,1 P. Puska,2 M. Arstila, P. Immonen-Räihä, E. Kaarsalo, M. Ketonen, S. Lehto, H. Miettinen, H. Mustaniemi, M. Niemelä, P. Palomäki, K. Pyörälä, M. Romo, V. Salomaa, J. Torppa, T. Vuorenmaa
National Institute of Health and Medical Research (INSERM U258) Paris
J. Richard,3 A. Bingham
National Institute of Health and Medical Research, Toulouse
J. Ferrières,1 J.B. Ruidavets, P. Marqués-Vidal, J.P. Cambou,2 P. Rodier, C. Saulet
Department of Epidemiology and Public Health–Faculty of Medicine, Strasbourg
D. Arveiler,1 P. Schaffer,1 A. Facello, E. MarineBarjoan, E. Uettwiller, D. Jacques
Pasteur Institute and Study and Research Group on Myocardial Infarction, Lille
P. Amouyel,1 M. Montaye-Faivre,1 J.-L. Salomez,2 M.-C. Nuttens,2 C. Graux, N. Marecaux
Bremen Institute for Prevention Research and Social Medicine
E. Greiser,1 B. Herman,4 G. Stüdemann
GSF-Institute for Epidemiology, Munich
U. Keil,1 H. Löwel, M. Lewis, A. Hörmann, J. Gostomzyk, H.D. Bolte
Center for Epidemiology and Health Research, Berlin
W. Barth,4 L. Heinemann,4 E. Classen, D. Quietzsch, G. Voigt, S. Brasche, S. Böthig
Heart Preventive Clinic, Reykjavik
N. Sigfusson,1 I.I. Gudmundsdottir, I. Stefansdottir, T. Thorsteinsson, H. Sigvaldason
National Institute of Health, Rome
A. Menotti,3 S. Giampaoli, A. Verdecchia
Institute of Cardiology, Regional Hospital, Udine
G.A. Feruglio,1† D. Vanuzzo, L. Pilotto, A. Antonini-Canterin, M. Scarpa, G.B. Cignacco, M. Spanghero, M. Palmieri
Research Center on Chronic Degenerative Diseases of the University of Milan
G.C. Cesana,1 M. Ferrario,1 G. De Vito, C. Bravi, M.T. Gussoni, R. Sega, R. Zanettini, O. Agostoni
Kaunas Medical Academy, Institute of Cardiology
J. Bluzhas,1 V. Grinius, R. Grazuleviciene, D. Rasteniene, D. Rastenyte
University of Auckland
R. Beaglehole,1 R.T. Jackson, A.W. Stewart
National Institute of Cardiology, Warsaw, Department of Cardiovascular Epidemiology and Prevention
S. Rywik,1 G. Broda, P. Kurjata, H. Wagrowska, M. Palakowska, W. Kupsc
National Research Center for Preventive Medicine, Moscow
T. Varlamova,1 V. Naumova, M. Ossokina, N. Serdyuchenko, N. Popova, E Bolshakova
Department of Health and Social Security, Barcelona
S. Sans,1 A. Puigdefàbregas, G. Paluzie, I. Balaguer-Vintró2
Department of Internal Medicine, Kalix Lasarett
F. Huhtasaari,1 V. Lundberg
Umeå University Hospital, Department of Medicine
The Queen’s University of Belfast, Northern Ireland
A.E. Evans,1 Z.M. Mathewson, E.E. McCrum, A. Hall, T. Falconer, E.L. McIlmoyle
University of Dundee, Scotland
H. Tunstall-Pedoe,1 C. Brown, M. Shewry
Royal Infirmary, Glasgow, Scotland
C. Morrison,4 G. Watt,5 W. Leslie, B. Fitzpatrick
Stanford Center for Research in Disease Prevention, Stanford, Calif
S.P. Fortmann,1 M. Winkleby, A. Varady, D. Jatulis, M. Hull
Novi Sad Health Center
M. Planojevi,1 D. Jakovlevic,2 T. Dapi, D. Stojšic, M. Krco, Z. Šolak
MONICA Management Center−World Health Organization, Geneva
I.Martin,6 I. Gyarfas,7 S. Böthig,7 Z. Pisa,7 S.R.A. Dodu,7 M.J. Watson, M. Hill
MONICA Data Center–National Public Health Institute, Helsinki, Finland
K. Kuulasmaa,6 J. Tuomilehto,7 A.-M. Rajakangas, E. Ruokokoski, V. Moltchanov, J. Torppa, M. Mähönen
MONICA Quality Control Center for ECG Coding–Hungarian Institute of Cardiology, Budapest
P. Ofner,6 A. Madai7
MONICA Quality Control Center for Event Registration, University of Dundee, Scotland
H. Tunstall-Pedoe,6 K. Barrett, C. Brown
MONICA Steering Committee
A. Evans (Chair), M. Hobbs (Chair, Publications Subcommittee), M. Ferrario, H. Tunstall-Pedoe (Rapporteur), K. Kuulasmaa, A. Shatchkute, (WHO, Copenhagen), I. Martin (WHO HQ, Geneva)
Consultants: A. Dobson, Z. Pisa, O.D. Williams
Previous Members of Steering Committee: S. Sans, F. Gutzwiller, R. Beaglehole, U. Keil, S.L. Rywik, S.P. Fortman, P. Puska, A. Menotti. Former Chiefs of CVD/HQ: V. Zaitsev (WHO, Copenhagen), J. Tuomilehto
Former Consultants: M.J. Karvonen (Helsinki, Finland), R.J. Prineas (Minneapolis, Minn); M. Feinleib (Bethesda, Md); F.H. Epstein† (Zürich, Switzerland)
2Former Principal Investigator
5Former Co-Principal Investigator
7Former Responsible Officer
MONICA Centers are funded predominantly by regional and national governments, research councils, and research charities. Coordination is the responsibility of the World Health Organization (WHO), assisted by local fund raising for congresses and workshops. WHO also contributes to the MDC in Helsinki, which is supported by the National Public Health Institute of Finland. A contribution to WHO from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md, provides some support for the MDC and the Quality Control Center for Event Registration in Dundee. Grants from ASTRA Hässle AB, Sweden, Hoechst AG, Germany, Hoffmann-La Roche AG, Switzerland, and the Institut de Recherches Internationales Servier (IRIS), France, support data analysis and preparation of publications. The authors would like to thank Esa Ruokokoski at the MDC and Marianna Chambless in Münster/Chapel Hill for their work in the analysis of these data and Carmen Ewe, Münster, for preparing the manuscript.
↵1 See “Appendix” for list of participants.
- Received June 30, 1997.
- Revision received August 22, 1997.
- Accepted August 29, 1997.
- Copyright © 1997 by American Heart Association
Löwel H, Dobson A, Keil U, Herman B, Hobbs MST, Stewart A, Arstila M, Miettinen H, Mustaniemi H, Tuomilehto J. Coronary heart disease case fatality in four countries. Circulation. 1993;88:2524–2531.
WHO MONICA Project, by Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project: registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994;90:583–612.
WHO MONICA Project, MONICA Manual. Revised ed. Geneva, Switzerland: Cardiovascular Diseases Unit, World Health Organization; 1990.
Sonke GS, Beaglehole R, Stewart AW, Jackson R, Stewart FM. Sex differences in case fatality before and after admission to hospital after acute cardiac events: analysis of community based coronary heart disease register. BMJ. 1996;313:853–855.
Vaccarino V, Krumholz HM, Berkman LF, Horwitz RI. Sex differences in mortality after myocardial infarction: is there evidence for an increased risk for women? Circulation. 1995;91:1861–1871.
Keil U, Chambless L. Sex differences in mortality after myocardial infarction: is there evidence for an increased risk for women? Circulation. 1995;92:3576–3577. Letter.
Tunstall-Pedoe H, Morrison C, Woodward M, Fitzpatrick B, Watt G. Sex differences in myocardial infarction and coronary deaths in the Scottish MONICA population of Glasgow 1985–91: presentation, diagnosis, treatment, and 28-day case fatality of 3991 events in men and women. Circulation. 1996;93:1981–1992.
Doll R, Muir C, Waterhouse J. Cancer Incidence in Five Continents. Berlin, Germany: Springer; 1970.
WHO Regional Office for Europe. Myocardial Infarction Community Registers. Results of a WHO International Collaborative Study, coordinated by Regional Office for Europe. Public Health in Europe No. 5. Copenhagen: 1976.
Sahasakul Y, Chaithiraphan S, Panchavinnin P, Jootar P, Thongtang V, Srivanasont N, Charoenchob N, Kangkagate C. Multivariate analysis in the prediction of death in hospital after acute myocardial infarction. Br Heart J. 1990;64:182–185.
Nidorf SM, Parsons RW, Thompson PL, Jamrozik KD, Hobbs MST. Reduced risk of death at 28 days in patients taking a β blocker before admission to hospital with myocardial infarction. BMJ. 1990;300:71–74.
McGovern PG, Folsom AR, Sprafka JM, Burke GL, Doliszny KM, Demirovic J, Naylor JD, Blackburn H. Trends in survival of hospitalized myocardial infarction patients between 1970 and 1985: the Minnesota Heart Survey. Circulation. 1992;85:172–179.