OBJECTIVE: To determine the completeness and acceptability of the cause of (COD) statements in death certificates before and after the conduct of an educational intervention aimed at improving COD certification in Province X.

METHODS: A random sample of 7939 death certificates from the year 1986-2006 was obtained. Part 1 of the COD section of the certificates was evaluated for completeness and for acceptability of the COD. Interrupted time series study was employed with semi-annual as the time point used in the analysis. The study covered 42 time points from 1986-2006. The proportions of complete and acceptable COD for each time point were calculated and segmented linear regression was applied to compare the time trend before and after the educational intervention.

RESULTS: Of the 7939 death certificates, only 35.3% were completely filled up, of which 16.5% had acceptable overall COD. The average proportions of all the death certificates with acceptable item entries in the immediate, antecedent, and underlying causes are 30.1%, 57.0% and 25.0%, respectively. The time series data on the acceptability of the COD statement revealed a significantly higher trend after the educational intervention.

CONCLUSION: A large proportion of the COD section of death certificates have incomplete and unacceptable entries. However, the acceptability of the COD statement improved after the educational intervention.