Depressive disorders during pregnancy have been shown to negatively affect mothers and infants in various ways, including poor obstetrical outcomes (1,2,3), higher risk of postpartum depression among mothers (4,5), low infant birth weight, and increased risk of premature delivery (2,6). Furthermore, some studies have found that a recent history of depression is related to women's noncompliance with scheduled prenatal care visits (7,8,9). Underutilization of prenatal care is detrimental to maternal and infant health (10,11,13).
Although there has been copious research reporting inadequate prenatal care among pregnant women with a recent history of depression (1,2,3,4,5,6,7,8,9), almost all of these studies adopted a survey design and retrospectively collected information in the form of women's self-reported past prenatal visits, which could entail misclassification. In addition, depressive symptoms are commonly associated with medical conditions and mental illnesses other than depressive disorders. Uncertainty regarding the relationship between depressive disorders and utilization of prenatal care stems from the heterogeneity of depressed populations identified by previous studies. Finally, nearly all previous studies focused on only the number or adequacy of prenatal care visits and rarely included costs of prenatal care in their analyses. Lacking this information, policy makers and clinicians may not put high priority on developing effective strategies to promote maternal and infant health in this group of women.
Therefore, with this study we aimed to use nationwide population-based data to compare medical utilization, including number of prenatal care visits and associated costs, between pregnant women who had received mental health care for a depressive disorder in the year before conception and those who had not. We hypothesized that pregnant women who had received mental health care for a depressive disorder would have fewer prenatal care visits and lower associated costs than the others.
This study used a database released by the Taiwan National Health Research Institute (NHRI) in 2008. This database includes 1,073,891 randomly selected participants, about 5% of the entire population of enrollees in the National Health Insurance (NHI) program. The database was created by the NHRI with the use of a systematic sampling method to form a representative sample from the entire population in Taiwan; thus there were no statistically significant differences in age, gender, and costs between the sample group and all enrollees. The database consists of medical claims for ambulatory care, inpatient care, dental services, and prescription drugs, as well as the registration for benefits of the insured from 1996 to 2006.
Because the database used in this study consists of deidentified secondary data released to the public for research purposes, this study was exempt from full review by the Internal Review Board.
We selected from the database all records of live births in Taiwan between January 1, 2004, and December 31, 2006 (N=23,591), excluding 288 multiple-birth deliveries (ICD-9-CM code 651.XX). We also excluded three live births to mothers who had any type of mental disorder (ICD-9-CM code 290.XX–319.XX) except depressive disorders (ICD-9-CM codes 296.2, 296.3, 300.4, and 311) within five years before delivery. In addition, we excluded ten pregnant women who received treatment for a depressive disorder during their pregnancy. Ultimately, our sample was the mothers of 23,290 infants.
The medical history of each mother was traced back to identify whether she had visited ambulatory care departments or had been hospitalized for treatment of a depressive disorder in the year before conception. We identified 614 participants (2.6%) who had received mental health care for depressive disorders in the year before pregnancy but who did not receive any treatment for their disorders during their pregnancy.
The key dependent variables were the number of prenatal care visits and associated costs. In Taiwan, the NHI provides free prenatal care and recommends ten prenatal visits for all pregnant Taiwanese women in order to reduce the risk of poor pregnancy outcomes and to decrease the need for pediatric care after birth. The key independent variable was whether a pregnant woman had received mental health care for a depressive disorder within one year before conception. Because the gestational age and delivery date of each infant are available in the data set, we were able to calculate the time of conception for each pregnant woman selected.
Bivariate analyses were conducted with Student's t tests and one-way analysis of variance. Poisson regression analysis was conducted to explore independent predictors of prenatal care visits. In addition, a multiple regression analysis was used to model the prenatal care costs as a linear function of a set of independent variables. The independent variables were sociodemographic factors (age and monthly income), chronic medical conditions (arterial hypertension, ICD-9-CM codes 642.0, 642.1, 642.2, 642.3, 642.9, and 760.0; diabetes, codes 648.0, 648.8, and 775.0; anemia, code 648.2; and cardiac disease, code 648.5 and 648.6), and obstetrical complications (malpresentation, ICD-9-CM codes 652, 761.7, 763.0, and 763.1; insufficient or excessive fetal growth, codes 656.5 and 656.6; and placenta previa or placental abruption, codes 641, 762.0, and 762.1).
A level of .05 was selected to determine the significance of predictors in the models. The SAS statistical package (SAS System for Windows, version 8.2) was used to perform the analyses in this study.
Women who had received mental health care for depressive disorders within the one-year period before conception were significantly more likely to have lower monthly incomes (p<.001), hypertension (p=.001), malpresentation (p=.039), and placenta previa or placental abruption (p<.001) than women who did not receive mental health care for depressive disorders. [Demographic characteristics and medical disorders for pregnant women in Taiwan are presented in tables available as an online supplement to this brief report at ps.psychiatryonline.org.]
Table 1 shows that mothers who received mental health care for depressive disorders had significantly fewer prenatal care visits (p<.001); the mean number of prenatal care visits was 8.50 and 9.17 for women who received and did not receive outpatient care for depressive disorders, respectively. Table 1 also provides the details of mean prenatal care costs for the sampled pregnant women. The cost of prenatal care visits for women who received care for depressive disorders were almost twice as high as for women without a history of depressive disorders ($NT 51,187 versus $NT 27,998; p<.001) and were also higher than the costs for women with several common chronic medical conditions and obstetrical complications, with the exception of anemia.
After we adjusted for women's age, monthly income, comorbid medical conditions, and obstetric complications, we found that mothers with a prior history of mental health service use for depression within one year before conception were significantly less likely to receive prenatal care (relative risk=.94, 95% confidence interval=.92–.97, p<.001) and about a $NT 22,494 greater cost (p<.001) than mothers who had not accessed mental health services for depression.
Further details of regression models can be found in tables available as an online supplement to this brief report at ps.psychiatryonline.org.
Our study found that pregnant women who received mental health care for a depressive disorder in the year before conception had significantly fewer prenatal care visits than other women but higher associated treatment costs. Our findings supported our hypothesis that women who received mental health care for a depressive disorder would have fewer prenatal care visits. However, the evidence contradicted our assumption that depressed women would therefore have lower prenatal care costs than others.
This finding agrees, in part, with the conclusions of previous studies that reported inadequate prenatal care among women with psychiatric histories (7,8,9). In Taiwan, with the inauguration of the NHI program in 1995, all pregnant women are allowed to have ten scheduled prenatal care visits for free. In other words, the NHI has eliminated financial burden as an obstacle to access to maternal health services. Despite reduced barriers to access, this study found that women who had received treatment for depression before conception tended to have fewer prenatal visits than other women.
Although pregnant women who received mental health care for depressive disorders had fewer prenatal care visits, they incurred greater prenatal care costs. This observation held after women's age, monthly income, medical conditions, and obstetric complications were adjusted for in the analyses. After reanalyzing our data we found that the higher costs were associated with prenatal treatment and drug costs. Although it was not possible to identify specific treatments or drugs that accounted for the higher cost, it might be related to obstetric complications or detrimental health behaviors aggravated by untreated depressive disorders, such as poor eating, poor sleeping, lack of exercise, and more smoking and drinking. Future studies should examine the relationship between obstetric complications, detrimental health behaviors, and higher prenatal care costs among pregnant women with a prior depressive disorder.
Limitations of this study should be noted. Prior studies have reported that some demographic risk factors, including education attainment, marital status, employment status, and number of previous deliveries, are associated with the number of prenatal care visits (7,14). However, the data set adopted in this study does not contain such information. The absence of adjustment for these factors should have little impact on interpretation of the findings, however, because we adjusted for age and monthly income—which can be proxies for education and employment status—in the regression modeling.
Another limitation is that the diagnoses of depressive disorders in the administrative database may be less accurate than diagnoses made according to a standardized diagnostic interview. However, the claims data were reported from hospitals, and they are considered to be relatively valid compared with those reported directly by patients in surveys used in prior studies. In addition, the misdiagnosis, if any, is likely to be nondifferential, which could only bias our results toward the null.
In addition, the rate of depression was low in this study (only 2.6% of pregnant women received mental health care in the preconception year). Moreover, less than 1% were treated for depression during pregnancy. Some women's depression may go undiagnosed. However, using a longer period, we found that up to 4.1% (947 out of 23,290) of the women had received mental health care for a depressive disorder in the five years before their pregnancy. This rate is similar to that reported by a review of 30 studies, which concluded that the prevalence of major depression ranged from 3.1% to 4.9% at different times during pregnancy (15). If we use this larger group of mothers as our study sample, the results remain quite similar.
Despite these limitations, this empirical study demonstrated that mothers who had received mental health care for depressive disorders before conception had fewer prenatal care visits and higher prenatal care costs than other pregnant women. To effectively promote maternal and infant health, policy makers cannot rely solely on offering ten prenatal care visits free of charge from medical institutions contracted under the NHI program. Health authorities should promote screening to identify pregnant women with a history of depressive disorders and provide mental health care to them in order to increase compliance with standardized prenatal visits. In addition, treatment for depression—ranging from antidepressant treatment to individual and group psychotherapy—might be prescribed to help them deal with emotional distress and encourage a healthy lifestyle. Such efforts may reduce prenatal care costs resulting from additional medical treatment needed for pregnant women who received inadequate prenatal care.
This study was supported by grant NSC 96-2314-B-227-005-MY2 from the National Scientific Council, Taiwan. This study was based on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health, Taiwan, and managed by the National Health Research Institutes. The interpretations and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institutes.
The authors report no competing interests.