Assessing Trends and Influencing Factors of Medical Services Utilization and Expenditures and the Likelihood of Employment Status Change among Women with Postpartum Depression—A Ten-Year Retrospective Matched Case-Control Study with National Claims

Project: A - Government Institutionb - Ministry of Science and Technology

Project Details


Postpartum depression (PPD) happens frequently to mothers, but is inadequately recognized and thus undertreated. To the present time, although the importance of PPD in women has been assessed in terms of its adverse effects on maternal health, it has not been methodically delineated among women in Taiwan, and the subject of risk of employment status change has rarely been evaluated, particularly. Hence, this study sets out to fill the gap. Specifically speaking, the two-year research aims to probe the following questions: (1). Measuring trends in incidence of women with PPD in Taiwan and evaluating patterns and influencing factors of medical services utilization and expenditures of the women; and (2). assessing the likelihood of employment status change (changing from a steady job to an unsteady one which implies the negative impact on economic productivity) among those women. Research methods and procedures of the present project are illustrated as follows: The First Year: (a). The research questions are to conduct a ten-year (year 2004 through year 2013) trend analysis of incidence of PPD among Taiwanese women, and analyze patterns and influencing factors of medical services utilization and expenditures of the women. (b). Study subjects of the research are women with PPD in Taiwan and will be retrieved from the Longitudinal Health Insurance Database 2010 (LHID2010) for the aforementioned ten-year study period. Since PPD is not given a distinct diagnostic entity in the ICD-9-CM, women with PPD will be identified based on other kinds of clinical depression. The LHID2010 will be linked with the Registration files. Subjects will then be aggregated by hospital-year as to produce a panel dataset. (c). A retrospective matched case-control study design will be employed. (d). The above-mentioned research questions will be firstly probed by conducting trend analysis, attempting to spot an underlying pattern. Age-specific rates for PPD will then be calculated for each year from 2004 through 2013 by region. In addition, Joinpoint regression analysis will be employed to analyze the trends in the age-adjusted PPD rates and determine if those changes are statistically significant. Lastly, Student’s t-test, chi-squared test or Fisher’s exact test, Mann-Whitney U test, nonparametric bootstrap methods, generalized estimating equation and propensity score methods will be conducted to evaluate predictors of medical services utilization and expenditures of the women. The Second Year: (a). The research objective is to assess the likelihood of changing employment status from income earners to non-income earners among those women during the ten-year period. (b). The aforementioned retrospective matched case-control study design and the panel dataset will be continuously used in the second year. (c). Statistical methods will be performed including chi-squared test, log-rank test, Kaplan-Meier survival analysis and multivariable Cox proportional hazards models. Given considerable medical and personal costs associated with PPD, increased attentions and efforts from policymakers, clinical practitioners and hospital administrators upon women with PPD are clearly warranted. It is hoped that the study could identify areas where investment might help reduce the incidence of PPD and the negative impact on economic productivity of the women, and thus make valuable contributions to improving the quality of life among pregnant women in Taiwan.
Effective start/end date8/1/147/31/15


  • Postpartum depression
  • Medical services utilization
  • Change in employment status


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.