WWW-publications from the WHO MONICA Project
October 20001
Markku Mähönen2, Hanna Tolonen2 and Kari Kuulasmaa2 for the WHO MONICA Project3
1 Columns "No. of events" of Table 10 were corrected on
23 January 2003.
The data for SPA-CATa were included in the published version of the Data Book on 19 December 2004.
2 MONICA Data Centre, National Public Health Institute, Helsinki, Finland
3 Annex: Sites and key personnel of the WHO MONICA
Project
This document updates the unpublished reports:
The MONICA Centres 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 supports the MONICA Data Centre (MDC) in Helsinki. Not covered by this general description is the ongoing generous support of the MDC by the National Public Health Institute of Finland, and a contribution to WHO from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland, USA for support of the MDC and the Quality Control Centre for Event Registration in Dundee. The completion of the MONICA Project is generously assisted through a Concerted Action Grant from the European Community. Likewise appreciated are grants from ASTRA Hässle AB, Sweden, Hoechst AG, Germany, Hoffmann-La Roche AG, Switzerland, the Institut de Recherches Internationales Servier (IRIS), France, and Merck & Co. Inc., New Jersey, USA, to support data analysis and preparation of publications.
This data book provides detailed descriptive statistics for each MONICA population on the data of the MONICA coronary event registers for the years 1980-1995. The data book consists of summary tables of the most important items for age group 35-64.
The report is based on the data which the MONICA Data Centre (MDC) has received from the MONICA Collaborating Centres (MCCs) on coronary events (Form 01), population statistics (Form A) and mortality statistics (Forms C and E).
In the specifications of the calculations for this data book the names of the data items of the Core Data Transfer Format-Coronary Events (1) have been used. The terminology used is this report follows that developed for MONICA event registration in the MONICA Manual (1), with later refinements in the collaborative publications (2, 3).
The report considers the Reporting Unit Aggregates (RUA) which are foreseen as potential candidates for units of analysis of the MONICA coronary event data. The RUAs, their abbreviations and Reporting Units are listed in Table 1. Some of the RUAs have several versions distinguished by suffix ´a´ and ´b´. Different combinations of Reporting Units (RU) may be used for analyses concerning coronary events only and for analyses involving both coronary event and risk factor data. The reason for the distinction is that some RUs of some RUAs were not included in every risk factor survey. Therefore, the different RUAs from AUS-PER, GER-BER, GER-BRE, GER-EGE, ICE-ICE, RUS-MOI and RUS-NOC may include same RUs. Furthermore, the combination (GER-AUG) of GER-AUR and GER-AUU has also been considered. Altogether 49 RUAs are considered.
The calendar years included in this report from each population are shown in Table 1.
Individual records have not been included in the analysis if:
Otherwise, all data available in the MDC were used in the analysis. The number of events excluded from the analysis, and the reasons for exclusions, are shown in Table 2.
Standardized methods for data collection were defined for the MONICA Project, and particular attention was paid on training of the personnel involved in the data collection, in data processing and in quality control. The quality of the event registration data and the demographic data are reported in separate quality assessment reports (4,5) No data were excluded from this book because of problems in the quality. Therefore the quality assessment reports should be consulted before using the data presented.
The definitions for diagnostic categories are shown in Appendix 1.
Diagnostic categories:
Management:
Necropsy:
Survival status:
Events were also classified as first events (PREMI= 6 or 7), recurrent events (PREMI=1, 2, 3 or 5) or indeterminate events (PREMI= 4 or 9). First events, recurrent events or indeterminate events are considered in tables and text only when specifically mentioned. Otherwise, all events are noted regardless of the previous history.
For many analyses, coronary events in several diagnostic categories were combined. The following definitions for a coronary event were used:
The age standardized event rates were calculated using the World standard population, with the following weights:
| Age group | 35-39 | 40-44 | 45-49 | 50-54 | 55-59 | 60-64 |
| Weight | 6 | 6 | 6 | 5 | 4 | 4 |
When reporting first events only, the year in which the availability of data on previous MI was less than 70% were not included.
The 95% confidence intervals for event rates were calculated using the relationship between the Poisson and chi-squared approximation to define the confidence intervals for weighted sums of Poisson parameters (6).
Case fatality was calculated as the proportion of fatal events in fatal and non-fatal events. Case fatality for the age group 35-64 was age-standardized using the following weights:
| Age group | 35-44 | 45-54 | 55-64 |
| Weight | 1 | 3 | 7 |
These weights were derived by looking at the overall number of fatal and non-fatal coronary and stroke events in MONICA data in MDC. If the denominator of the youngest age group was zero, the two youngest age groups were combined, using weight 4. (This occurred occasionally for women.) If the denominator of one of the older age groups was zero, no age standardization was used. (This did not occur in the current data book.) For case-fatality the 95% confidence intervals were calculated in the conventional manner using the normal approximation of the binomial distribution (7). The 10-year age specific case fatality was only printed in the tables if the number of fatal events was at least 10.
When reporting first events only, the years in which the availability of data on previous MI was less than 70% were not included.
Table 2 presents the number of events included in or excluded from the other tables of the data book. Note that women were not registered in SWI-TIC and SWI-VAF, and therefore the data book does not show the results for women in these RUAs.
Table 3.1 presents the proportions of first, recurrent and indeterminate events among fatal events (F1+F2+F9). In men, the proportion of indeterminate events was below 20% over all years in CHN-BEI, FIN-KUO, FIN-NKA, FIN-TUL, GER-AUR, ICE-ICEa, ICE-ICEb, ITA-BRI, ITA-FRI, NEZ-AUC, RUS-NOCa, RUS-NOCb, RUS-NOIa, SWE-GOT, SWE-NSW, UNK-BEL, UNK-GLA and YUG-NOS. The proportion was between 20-30% in one or several years (but not over 30%) in FRA-TOU, DEN-GLO, LTU-KAU and USA-STA. In other RUAs the proportion was over 30% in one or several years, which probably makes the data unsuitable for analyses in which first event data are needed. In women, there was more variation because of small numbers.
Table 3.2 presents the proportions of first, recurrent and indeterminate events among non-fatal definite MI events (NF1). The proportion of indeterminate events was generally very low, but it was high in CAN-HAL, HUN-BUD, the first year in HUN-PEC, and somewhat higher in the first study years in DEN-GLO and USA-STA and in 1988 in GER-EGE.
Table 4.1 presents fatal (F1+F2+F9) events by different categories of management. There was marked variation in the proportion of hospitalized events. There was more variation in women. In men the proportion of hospitalized events was high in DEN-GLO in 1982-1983, in FRA-STR in 1984, in GER-RHN in 1984 and in POL-WAR in 1984. It was low in in CHN-BEI in 1993, in GER-AUR in 1994, in GER-BREa in 1992, in POL-TAR during the whole study period, and very low in SWE-GOT in 1992-1994. In SWI-TIC and SWI-VAF the availability of data for coding of MANAGE was low and varied over time (see Section 10 of the quality assessment report for specific comments). The proportion of hospitalized fatal events thus varied between different RUAs between 5%-100%. There may have been a change in coding during the study period in DEN-GLO, POL-WAR and SWE-GOT. A special case is POL-TAR where only a small proportion of fatal events are coded as hospitalized. Overall, the proportion of fatal events coded as hospitalized declined during the study period. In some RUAs, a large proportion of fatal events were coded as out-of hospital deaths, but medically attended (other medically attended).
The differences observed in the proportion of fatal events coded as hospitalized may reflect differences in the management of patients or differences in the interpretation of coding rules for the item MANAGE. The possible differences in coding practices should be taken into account when calculating trends of in-hospital and out-of hospital mortality rates within a defined population and especially when comparing different populations.
Table 4.2 presents nonfatal definite MI (NF1) events by different categories of management. Almost all non-fatal events (99% or over during all years) have been coded as hospitalized in most RUAs. Other codes were used in men in about 2-3% of NF1 events in the following RUAs: BEL-CHA, BEL-GHE, CHN-BEI, FIN-TUL, HUN-BUD, ICE-ICEb, LTU-KAU, RUS-NOCa, RUS-NOCb, RUS-NOI, SWE-NSW and UNK-GLA. In women proportionally more events were coded as non-hospitalized, but the numbers were small.
Table 5 shows the proportion of autopsies in fatal events (F1+F2+F9). There was a large variation in the frequency of autopsies between the RUAs. However, within each RUA the frequency of autopsies was fairly stable except in DEN-GLO, GER-EGEa, GER-EGEb and SWE-GOT where there was a drop in the frequency of autopsies, and in YUG-NOS where there was first a rise in the frequency of autopsies and later a marked decline. The frequency of autopsies was quite low in BEL-CHA, BEL-GHE, BEL-LUX, CHN-BEI, FRA-LIL, FRA-TOU, GER-AUR, GER-AUU, GER-BRE, POL-TAR, POL-WAR, SPA-CAT, SWI-TIC, SWI-VAF.
Table 6 presents the event rates for all events, including both first and recurrent events. The number of NF1 events is presented in the left-hand column and to the right event rates using different definitions of MI. The event rates differ notably between the different definitions. The event rates for years 93-95 in YUG-NOS are not reported because the demographic data were not available at the time when this data book was prepared.
Table 7 presents the event rates for first coronary events in the same format as used for event rates for all events. If the availability of data on whether the event was first or recurrent was less than 70% the event rate was not calculated. There are several RUAs for which event (incidence) rates cannot be calculated reliably.
Table 8 and Table 9 present the case fatality (CF) rates for all events and first events, respectively. CF is presented as 28 day CF and as < 24 hour CF. The tables show that almost all the deaths occur within 24 hours, and in many RUAs but not all the 24 hours CF seems to be declining during the last years. Table 10 presents the proportions of deaths by cumulative time to death (%).
Table 11 presents age standardized death rates by MONICA categories (F1+F2+F9) and (F1+F2), and from official mortality statistics. In some RUAs the rates obtained from the official mortality statistics are quite close to the rates obtained by the MONICA category (F1+F2+F9); however, in several RUAs there were notable differences between the routine mortality statistics and MONICA categories.
Table 12 presents proportions of NF1 events in different subcategories specified by the diagnostic findings. Notable variation from year to year in the proportion may indicate changes in the quality of the individual data items (ECG, ENZ, SYMPT) or in the availability of data. Both can have an impact on the classification of events in diagnostic categories. The classification of events to diagnostic category 1 (NF1) was based on abnormal ECG (ECG=1) in over half of the cases in all RUAs. In some RUAs the difference between the highest and lowest proportion with abnormal ECG was over 40% (BEL-CHA; DEN-GLO; HUN-BUDa; HUN-PEC; RUS-NOI; SWE-GOT; YUG-NOS).
Table 13 shows the proportion of registered fatal events which were classified as definite MIs only on the basis of symptoms, ECG and enzymes, and the proportions of the remaining fatal events as they were classified on the basis of symptoms, previous history and competing diagnoses to diagnostic categories possible coronary death, unclassifiable death and other (no MI) death. After that the events were classified according to the results of autopsy (third row of the table). The table shows that there were notable differences between the RUAs in the way the autopsy changed the classification of the preliminary categories of possible coronary deaths and unclassifiable deaths.
One of the observations from Table 13 is that it is very rare that events otherwise diagnosed as unclassifiable become coded as no MI after the necropsy. This observation, however, does not mean that nearly all of the unclassifiable deaths would be coded as definite or possible MI if necropsy had been done. The reason for the misleading observation is that it is based on events where the results of necropsy were available, and if the results of the necropsy had been negative, it would be unlikely that the event had met the criteria for being registered in MONICA.