Chart-Recorded Psychiatric Diagnoses in Women Giving Birth in California in 1992

The prevalence of psychiatric and substance use disorders is thought to be the same in pregnant women as in the general population of women (1–3). Few investigations, however, have assessed the reporting, diagnosis, or treatment of psychiatric disorders in the obstetrical sector. In one of the few population-based studies undertaken, Chasnoff et al. (3) identified an overall underreporting of substance abuse by obstetricians.

This paucity of information is potentially of concern, as the long-term sequelae of untreated maternal psychiatric and substance use disorders can be severe for both mother and infant (4, 5). Prenatal visits and hospital deliveries provide an opportunity to evaluate women for psychiatric disturbance and initiate treatment.

As an initial step toward understanding the detection of psychiatric disturbance among pregnant women, and to better describe the current reporting patterns by obstetrical providers, we undertook a population-based investigation of ICD-9 psychiatric and substance use diagnoses in all women giving birth in California in 1992. Our primary research aim was to assess the chart-recorded frequencies of psychiatric and substance use diagnoses in women at delivery.

METHOD

We undertook a secondary analysis using the California Health Information for Policy Project data set. The data set is composed of linked California vital statistics birth records and hospital discharge records (6). Further details on the linkage were presented by Herrchen et al. (6). The subjects for this study included all women (N=580,282) who delivered in a California civilian hospital in 1992.

Psychiatric and substance use diagnoses were defined by ICD-9 codes recorded on maternal discharge summaries. We examined the frequencies of psychiatric disorders using the following diagnostic categories: substance-related, schizophrenic, mood, anxiety, and other disorders. Two obstetrical ICD-9 diagnostic categories were also examined, as they can be used in lieu of more specific ICD-9 psychiatric diagnoses: “mental disorder of pregnancy” (ICD-9 648.4) and “substance use disorder of pregnancy” (ICD-9 648.3). The following disorders were excluded: disorders of childhood, delirium, dementia and other cognitive disorders, diagnoses “due to a general medical condition,” and personality disorders. In addition, we ascertained the frequency of women receiving psychiatric, substance use, and dual diagnoses.

RESULTS

The racial breakdown of the population was as follows: 36.5% Caucasian, 7.1% African American, 45.7% Hispanic, 6.8% Asian, 0.4% Native American, 3.5% other. Twenty-three percent of women were covered by private insurance, 26% by health maintenance organizations, 48% by Medi-Cal, and 3% had no insurance. The mean age was 26.9 years (SD=6.1), and 40% of mothers had one child. Sixty-six percent of women were married, and 64% had at least 12 years of education. The mean length of hospital stay for delivery was 2 days (SD=2).

Table 1 displays the frequencies of diagnoses recorded. Approximately three-fourths of the diagnoses were substance use disorders. The total number of women with psychiatric diagnoses was 8,828, suggesting that many women received more than one diagnosis. Of the 580,282 women who delivered, the frequency of any psychiatric or substance use disorder recorded was 1.5%. The frequency of psychiatric diagnoses alone was 0.2% (N=1,088), substance use diagnoses alone, 0.9% (N=5,217), and dual diagnoses, 0.4% (N=2,523).

DISCUSSION

This is the first population-based investigation to examine the reporting of psychiatric and substance use disorders among pregnant women in the general obstetrical sector. A review of epidemiological literature regarding psychiatric disturbances in pregnant women reveals prevalences between 9% and 18% for depression (1, 2) and between 10% and 20% for substance abuse (3). Thus, our data suggest that obstetrical practitioners throughout the state of California are not documenting psychiatric and substance use disorders on delivery discharge summaries for a substantial number of women.

Research into the detection and reporting of psychiatric disturbances among pregnant women is in its initial stages, making it difficult to draw definitive conclusions from these findings. During the peripartum period, obstetrical practitioners may be reluctant to diagnose psychiatric disorders, as this period can encompass a range of acute emotional and behavioral responses and may not be a reliable indicator of ongoing psychiatric conditions.

Of most concern is the possibility that this represents an underrecognition of ongoing antepartum disturbance. Prior investigations have documented that nonrecognition of affective disorders by primary care providers leads to undertreatment, which has deleterious effects on psychological and functional outcomes (7). An additional concern in the obstetrical sector is that underrecognition and undertreatment of disorders during pregnancy potentially affect both maternal and infant outcomes, including pregnancy and delivery complications (8), postpartum psychiatric disturbances (2), adverse developmental consequences for the infant (4), and future refractory maternal illnesses (5).

A potentially less serious explanation is that the low frequency of chart-recorded diagnoses reflects an underreporting rather than an underrecognition of psychiatric disorders. This implies that obstetrical practitioners are recognizing, diagnosing, and/or treating maternal psychiatric illness but not recording diagnoses. However, recording diagnoses at discharge in both maternal and infant charts serves to alert current and future medical providers (including pediatricians) and may help provide appropriate treatment.

To our knowledge, this investigation is the first to report statewide obstetrical documentation patterns for psychiatric and substance use illness. Obstetricians used generic obstetrical codes to categorize psychiatric disturbance for over 40% of diagnoses. As suggested in primary care medicine, nonpsychiatric practitioners may find psychiatric codes cumbersome to use, may have deficiencies in knowledge in diagnosing psychiatric disorders, or may fear that patients will be stigmatized if given psychiatric diagnoses (9). Furthermore, obstetricians more frequently documented substance use disorders, which may in part be explained by the availability of urine toxicology screens as a diagnostic tool. The high proportion of comorbid diagnoses recorded may signify that obstetricians are detecting women with severe and complicated symptoms.

It is important to consider that these results do not represent prevalence estimates of psychiatric disturbance at delivery; they represent only the reporting of disorders by obstetricians. Furthermore, we were unable to assess the validity of the recorded psychiatric diagnoses.

Although this investigation is an important initial step, further research into the recognition and treatment of psychiatric disturbance during pregnancy and at delivery is needed. Future investigations will need to employ chart review and structured interviews, as well as to assess the impact of educational programs or diagnostic screens to improve obstetrical detection. Ultimately, it is important to evaluate the effect of improved recognition on maternal pregnancy outcome and the health and development of the infant.

Presented in part at the Young Investigators Colloquium at the 150th annual meeting of the American Psychiatric Association, San Diego, May 17–22, 1997. Received March 30, 1998; revisions received Aug. 10 and Oct. 13, 1998; accepted Nov. 19, 1998. From the Departments of Psychiatry, Epidemiology and Preventive Medicine, and Obstetrics and Gynecology, University of California, Davis; Health Information Solutions, Redwood City, Calif.; and the Department of Psychiatry, University of California, Los Angeles. Address reprint requests to Dr. Kelly, University of California, Davis Medical Center, Department of Psychiatry, 2230 Stockton Blvd., Sacramento, CA 95817; [email protected] (e-mail). Supported by NIMH National Research Service Award MH-11892 to Dr. Kelly and the University of California, Davis. The California Health Information for Policy Project data set is supported by a grant from the Robert Wood Johnson Foundation.

TABLE 1

Enlarge table