Quality Assessment of Coronary Event Registration Data in the WHO MONICA Project
This appendix describes a scoring system developed for 10-year trends in coronary events. The Reporting Unit Aggregates (RUAs) and years considered are specified in Table A1.
A summary quality score is compiled from five component scores:
and other knowledge of major quality problems.The scores are defined below, and their values for each RUA are shown in:
Item 1. Unresolved constraint violations (Table 4 of coronary event quality assessment):
| 2 | if less than 10 in every year |
| 1 | otherwise |
Item 2. Unresolved suspected duplicates (Table 5 of the coronary event quality assessment):
| 2 | if less than 3 in every year |
| 1 | otherwise |
General items score: 0.1 × (Item 1 + Item2)/2
Item 3. The coverage of fatal registration (Table 6 of the coronary event quality assessment):
| 2 | if CHD/CHD ratio more than or equal to 1.0 in every year except possibly one year
(three or less missing events are accepted for each year, but even the possible
exceptional year must have the ratio more than or equal to 0.9); AND missing or erroneous routine mortality statistics for no more than one year. |
|
| 1 | if the criteria for code 2 are not met but the CHD/CHD ratio is more than 0.85 in every year. | |
| 0 | otherwise. |
Fatal coverage score: 0.3 × Item 3.
Item 4. The differences in trends in event rates in men between Definition 1 and Definition 2 (Table A2):
| 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 5. The differences in trends in case fatality in men between Definition 1 and Definition 2 (Table A2):
| 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 6. The differences in trends in event rates in women between Definition 1 and Definition 2 (Table A2):
| 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 7. The differences in trends in case fatality in women between Definition 1 and Definition 2 (Table A2):
| 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: 0.3 × (Item 4+Item 5+Item 6+Item 7)/4, except in SWI-TIC and SWI-VAF where no data was collected for women (0.3 × (Item 4 + Item 5)/2).
Item 8. Availability of >=3 ECGs (Table 8 of the quality assessment report):
| 2 | if more than or equal to 80% in every year |
| 1 | if less than 80% in one or several years but the availability of data on enzymes >= 90% in every year |
| 0 | otherwise |
Item 9. Availability of data on PREMI in nonfatal definite (NF1) events in men (databook table 3.2):
| 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 |
Item 10. Availability of data on PREMI in fatal (F1+F2+F9) events in men (databook table 3.1):
| 2 | if less than or equal to 10 % indeterminate events in every year, except possibly one year |
| 1 | if more than 10 % but less than or equal to 20% indeterminate events (one year over 20% accepted) |
| 0 | otherwise |
Item 11a. Trends in the proportion of NF1 from (NF1+NF2) in men (Table A2):
| 2 | if less than 1.5 % |
| 1 | if more than or equal to 1.5 % but less than 2.5 % |
| 0 | otherwise |
Item 11b. Trends in the proportion of NF1 from (NF1+NF2) in women (Table A2):
| 2 | if less than 1.5 % |
| 1 | if more than or equal to 1.5 % but less than 2.5 % |
| 0 | otherwise |
Item 12a. Trends in the proportion of ECG=1 from NF1 events in men (Table A2):
| 2 | if less than 1.5 % |
| 1 | if more than or equal to 1.5 % but less than 2.5 % |
| 0 | otherwise |
Item 12b. Trends in the proportion of ECG=1 from NF1 events in women (Table A2):
| 2 | if less than 1.5 % |
| 1 | if more than or equal to 1.5 % but less than 2.5 % |
| 0 | otherwise |
Item 13a. Trends in the proportion of ECG=1 and ENZYME different from 1 among the NF1 events in men (Table A2):
| 2 | if less than 1.0 % |
| 1 | if more than or equal to 1.0 % but less than 2.0 % |
| 0 | otherwise |
Item 13b. Trends in the proportion of ECG=1 and ENZYME different from 1 among the NF1 events in women (Table A2):
| 2 | if less than 1.0 % |
| 1 | if more than or equal to 1.0 % but less than 2.0 % |
| 0 | otherwise |
Other items score: 0.2 × (Item 8 + Item 9 + Item 10 + Item 11a + Item 11b + Item 12a + Item 12b + Item 13a + Item 13b)/9, except in SWI-TIC and SWI-VAF where no data was collected for women (0.2 × (Item 8 + Item 9 + Item 10 + 2×Item 11a + 2×Item 12a + 2×Item 13a)/9).
Item 14. The quality items on external quality assessment are derived from the proportions of agreement in "Major items" and "Other items" of the test case history coding in 1987 and 1989 (Appendix 3 of the quality assessment report) and in "MONICA code" of the ECG test coding in 1987 and 1989 (Appendix 4 of the quality assessment report). If data we not available on "MONICA code" the agreement was considered to be 0.7 in 1987 and 0.6 in 1989.
EQA score: 0.1 × [(Major items × Other items × MONICA code) in 1987 + (Major items × Other items × MONICA code) in 1989].
There were other major quality problems in the RUAs:
The problems are described in more detail in Section 10: Comments on individual RUAs of the quality assessment report.
The Summary Score is defined as the sum of the 5 component scores, except for the RUAs with other major quality concerns (i.e. AUS-PERb, FRA-TOUa, POL-TARa, RUS-MOCa, RUS-MOIb, RUS-NCb, RUS-NOIa, SWI-TICa and SWI-VAFa), for which 0.6 was subtracted from the sum. Before the subtraction, the contributions of each of the five component scores are: