Dengan menyadari sepenuhnya bahwa hasil penelitian ini tidak dapat digeneralisir untuk seluruh populasi yang ada, namun berdasarkan temuan hasil di lapangan, perlu kiranya dicermati dan diupayakan suatu bentuk intervensi metode pendidikan kesehatan yang lainnya. Sudah barang tentu akan lebih bermanfaat apabila langsung digali dari potensi daerah setempat serta dipertimbangkan kondisi situasionalnya.

---

It seems that the high maternal mortality rate in Indonesia is an interesting problem to be researched. The pregnancy maintenance or examination is very important in an effort to prevent a prenatal mortality and is also related with other components such as the infant mortality rate (as one of the public health indicators).This research was performed in the District of Blitar, for about three months (April-June 1996) with total responden 81 prep nant woman. The respondents were divided into 40 subjects as a treatment group anf the rest 41 subjects as a control group. The sample selection is done with purposive random sampling.The method used in this research is a Quasi Experimental (pre-post design), in which there are many factors which actually influenced the process was not accounted for or cannot be control led previously.The intervention given to the treatment group is a calendar of pregnancy testing which is designed by researcher by modifying the Pregnant Woman Card (KMS). Before the calendar were given to the pregnant woman, it was previously tested to other group which has relatively the same characteristics with the researcher respondents subjects.This media was tried an innovation, so this media is newly used in the researched region, that the respondent interest to used it is quite high.The testing is was done through univariate analysis test by showing the frequency distribution table of the respondent characteristics.The bivariate analysis so far has been used to obtain the a sufficient knowledge and understanding regarding the researched objects. Then the multivariate analysis is intended to understand the influence of independent variable toward its dependent variable in terms of the significance of the result of the statistic test used.The results of the research indicates that the calendar methods have quite significant influence toward the knowledge, confidence and attitudes variables (p < 0.05), its showed the increase of dependent variables. The results is that the increase of knowledge variable in the intervention group is influenced by the characteristics of the pregnant woman which includes the parity and occupation. While for other variables such as the confidence is influenced by the partus record, and then the attitudes variable is influenced more by education.We are fully convinced that this research results can not be used for all population.However,based on the findings and results in the field, it should be observed and tried to make other kind of intervention health education method. However, it will be far more beneficial if it is directly explored from the local area and based on the situational condition.