WWW-publications from the WHO MONICA Project
November 2001
Markku Mähönen1, Kjell Asplund2, Hanna Tolonen1 and Kari Kuulasmaa1 for the WHO MONICA Project3
1 Department of Epidemiology and Health Promotion, National Public Health
Institute (KTL), Helsinki, Finland
2 Department of Medicine, University Hospital, Umeå, Sweden
3 Annex: Sites and key personnel of the WHO MONICA
Project
Correspondence to: markku.mahonen@ktl.fi
The quality assessment of stroke event registration data in the WHO MONICA Project has been published [1]. The purpose of this document is to derive quality scores for the data for the analysis of trends in the rates of stroke events during the 10-year period of the Project. Such quality scores will be needed both when reporting the 10-year trends in stroke events and for testing the Project's first main stroke hypothesis. The hypothesis explores the relationship between major CVD risk factors of serum cholesterol, blood pressure and cigarette consumption and 10-year trends in incidence rate of stroke [2].
Data quality indicators chosen for the quality scores are more comprehensive than those discussed in an earlier evaluation of case ascertainment in the MONICA stroke study [3]. The earlier indicators have been supplemented with items about the reliability of the assessment of trends and items about general data quality. An indicator based on 28 day case fatality has been excluded because the observed differences may reflect true differences in case fatality rather than problems in data quality.
The MONICA Reporting Unit Aggregates (RUAs) and years considered are shown in Table 1. The report considers the RUAs which are foreseen as potential candidates for units of analysis of 10-year trends in stroke event rates using the data from the WHO MONICA Project. Compared with the quality assessment report [1], calendar years were excluded from the beginning or the end of the even registration period of some RUAs (CHN-BEI, RUS-NOC, RUS-NOI), because the exclusion improved markedly the quality of the data for the remaining period in these RUAs. RUAs where data were available for less than 7 years were not included.
The analyses were done using data for age group 35-64 years. Data for men and women were combined because no significant differences in the quality of the data could be found between the two sexes, and combining data sets for men and women increased the statistical precision of the findings.
The report is based on the findings of the quality assessment of the MONICA data [1] and the stroke event registration data book [4], as well as on direct calculations from the individuals' stroke event data which the MONICA Data Centre has received from the MONICA Collaborating Centres. There are minor discrepancies between the summary results reported here and the detailed tables of the quality assessment report [1] used as reference, because the quality assessment report was calculated for age group 25-64. Furthermore, for Items 11 and 13, the quality assessment tables include diagnostic category 4, whereas the calculations used here for the items do not.
The Stroke Event Trend Quality Score is compiled from five components:
In this section we first specify the calculations for each of the component scores and then give the values of the component scores for each RUA. The term "DIACAT" refers to the MONICA diagnostic category [5]:
| DIACAT | Diagnostic category |
|---|---|
| 1 | definite stroke |
| 5 | definite stroke associated with a definite coronary event |
| 9 | unclassifiable |
| 4 | not stroke |
This score summarizes the information from general data quality items related to data management and availability of data.
Item 1. Unresolved constraint violations (Table 4 of stroke event quality assessment [1]. No age limits were applied to this item.):
| 2 | if less than 10 in every year |
| 1 | otherwise |
Item 2. Unresolved suspected duplicate registrations (Table 5 of stroke event quality assessment [1]. No age limits were applied to this item.):
| 2 | if less than 3 in every year |
| 1 | otherwise |
Item 3. Availability of data on previous stroke in fatal events (DIACAT = 1 + 5 + 9) in men (Table 3.1 of stroke event data book [5]):
| 2 | if less than or equal to 10% indeterminate events in every year or every year except one |
| 1 | if more than 10% but less than or equal to 20% of indeterminate events (one year more than 20% accepted) |
| 0 | otherwise |
Item 4. Availability of data on previous stroke on nonfatal events (DIACAT = 1 + 5 + 9) in men (Table 3.2 of stroke event data book [5]):
| 2 | if less than or equal to 3% indeterminate events in every year |
| 1 | if more than 3% but less than 10% indeterminate events (one year over 10% accepted) |
| 0 | otherwise |
General Items Score: (Item 1 + Item 2 + Item 3 + Item 4)/4
This score reflects the coverage of fatal event registration by comparing it with the official mortality statistics. Even though all deaths officially coded as strokes are not necessarily strokes in the MONICA classification, all of them should have been registered for the MONICA validation.
Item 5. The coverage of fatal event registration in comparison with routine mortality statistics (Table 7 of stroke event quality assessment [1]):
| 2 | if STR/STR ratio more than or equal to 0.9 in every year and Total/STR ratio is more than or equal to 1.0 in every year |
| 1 | if the criteria of code 2 are not met but the Total/STR ratio is more than 0.95 in every year or every year except one |
| 0 | otherwise |
Fatal Coverage Score: Item 5.
Item 6. Trends in the proportion of events eventually diagnosed as not being strokes (DIACAT = 4) of all registered fatal events, men and women combined (Table 2):
| 2 | if less than 1.0% |
| 1 | if more than or equal to 1.0% but less than 2.0% |
| 0 | otherwise |
Item 7. Trends in the proportion of non-hospitalized fatal events of all fatal events (DIACAT = 1 + 5 + 9), men and women combined (Table 2):
| 2 | if less than 1.0% |
| 1 | if more than or equal to 1.0% but less than 2.0% |
| 0 | otherwise |
Item 8. Trends in the proportion of non-hospitalized nonfatal events of all registered nonfatal events, men and women combined (Table 2):
| 2 | if less than 1.0% |
| 1 | if more than or equal to 1.0% but less than 2.0% |
| 0 | otherwise |
Item 9. Trends in the proportion of events eventually diagnosed as not being strokes (DIACAT = 4) of all registered nonfatal events, men and women combined (Table 2):
| 2 | if less than 1.0% |
| 1 | if more than or equal to 1.0% or less than 2.0% |
| 0 | otherwise |
Item 10. Overall proportion of events eventually diagnosed as not being strokes (DIACAT = 4) of all registered fatal events (Table 8 of stroke event quality assessment [1]):
| 2 | if more than 10% in every year |
| 1 | if less than or equal to 10% but more than 5% in every year |
| 0 | otherwise |
Item 11. Overall proportion of non-hospitalized fatal events of all fatal events (DIACAT = 1 + 5 + 9) (Table 9 of stroke event quality assessment [1]):
| 2 | if more than 10% in every year |
| 1 | if less than or equal to 10% but more than 5% in every year |
| 0 | otherwise |
Item 12. Overall proportion of events eventually diagnosed as not being strokes (DIACAT = 4) of all registered nonfatal events (Table 10 of stroke event quality assessment [1]):
| 2 | if more than 10% in every year |
| 1 | if less than or equal to 10% but more than 5% in every year |
| 0 | otherwise |
Item 13. Overall proportion of non-hospitalized nonfatal events of all nonfatal events (DIACAT = 1 + 5 + 9) (Table 11 of stroke event quality assessment [1]):
| 2 | if more than 10% in every year |
| 1 | if less than or equal to 10% but more than 5% in every year |
| 0 | otherwise |
Other Important Coverage Items Score: (Item 6 + Item 7 + Item 8 + Item 9 + Item 10 + Item 11 + Item 12 + Item 13)/8
This score reflects the reliability of the data in the assessment of trends in event rates.
Item 14. The difference in trends in event rates between Definition 1 (i.e. DIACAT = 1, 5 or 9) and Definition 2 (i.e. DIACAT = 1 or 5), men and women combined (Table 3):
| 2 | if less than 0.5% |
| 1 | if more than or equal to 0.5% but less than 1.0% |
| 0 | if more than or equal to 1.0% |
Item 15. The difference in trends in mortality rates between Definition 1 and Definition 2, men and women combined (Table 3):
| 2 | if less than 0.5% |
| 1 | if more than or equal to 0.5% but less than 1.0% |
| 0 | if more than or equal to 1.0% |
Item 16. The difference in trends in case fatality between Definition 1 and Definition 2, men and women combined (Table 3):
| 2 | if less than 0.5% |
| 1 | if more than or equal to 0.5% but less than 1.0% |
| 0 | if more than or equal to 1.0% |
Trend Reliability Score: (Item 14 + Item 15 + Item 16)/3.
This score reflects the reliability of the diagnostic classification of the events.
Item 17. Trends in proportion of fatal events with no diagnostic information (no necropsy, not seen by a physician or neurologist before death) in all fatal events (DIACAT = 1 + 5 + 9), men and women combined (Table 2):
| 2 | if less than 1% |
| 1 | if more than or equal to 1% but less than 2% |
| 0 | otherwise |
Item 18. Overall proportion of fatal events with no diagnostic information in all fatal events (DIACAT = 1 + 5 + 9) (Table 13 of stroke event quality assessment [1]):
| 2 | if less than 5% in every year |
| 1 | if more than or equal to 5% but less than 20% in every year |
| 0 | otherwise |
Reliability of Classification Items Score: (Item 17 + Item 18)/2
The values of the quality items and component scores for each RUA are shown in :
The Stroke Event Trend Quality Score is defined as the weighted mean of the 5 component scores, such that the maximum score is 2. The five components are weighted as:
The values of the Stroke Event Trend Quality Score for each RUA are shown in Table 6.
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 was 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.