Gazmararian, Julie A. PhD; Petersen, Ruth MD; Jamieson, Denise J. MD; Schild, Laura; Adams, Melissa M. PhD; Deshpande, Anjali D. PhD; Franks, Adele L. MD
Many women experience morbidity during pregnancy, resulting in additional outpatient care, inpatient care while pregnant, pregnancy loss, extended hospitalization before delivery, and complications during labor and delivery. In 1992, for every 100 US hospitalizations involving a birth, an estimated 18.1 nondelivery hospitalizations occurred for complications: 13.7 for antenatal complications and 4.4 for pregnancy loss.1 Similar data have been found among state (25 per 100 deliveries),2 military (26.8% of women hospitalized),3 and clinic‐based (17.7% of women hospitalized)4 populations, as well as in other countries.5
The only large datasets available to monitor antenatal morbidity derive from administrative sources. Because these sources usually cannot distinguish routine outpatient antenatal care from such care associated with antenatal morbidity, most antenatal morbidity measurements have used hospitalization as an indicator of severe morbidity. The National Hospital Discharge Survey has been the primary data source for monitoring antenatal hospitalizations.1,6 However, this database has several limitations, notably the inability to determine the number of multiple hospitalizations for one woman, identify prolonged hospitalizations that immediately precede delivery, and provide detailed information about admission diagnoses and associated costs. Additional data could be instrumental in understanding the burden of antenatal hospitalizations by providing information about extended hospitalizations or separate admissions before delivery, as well as the characteristics of these hospitalizations.
We used administrative data from a large managed care organization to examine the prevalence of hospitalizations during pregnancy, factors associated with hospitalizations during pregnancy, the distribution of the number of hospital visits during pregnancy, and the reasons, length of stay, and associated costs of inpatient care. This information may be useful to monitor trends in care and to assess the financial and service burdens associated with severe antenatal morbidity.
MATERIALS AND METHODS
The study population was drawn from women who were enrolled in a national commercial managed care plan and had a pregnancy‐related hospital discharge code from January 1 through December 31, 1997 (n = 70,543, Figure 1). Women who had insufficient information to determine pregnancy conclusion, had an invalid code (eg, age, sex), or were not continuously enrolled for the 308 days before their delivery or pregnancy loss date were excluded, leaving 46,179 eligible women. Of these, 35 had two pregnancies during the study period, and information from both pregnancies was included.
We first classified women as having either a live birth or pregnancy loss outcome (contact author for details). We then classified women with live births as being hospitalized during pregnancy if they had any inpatient claim record within 259 days before their delivery date, or up to 308 days before the delivery date if the hospitalization also included a pregnancy‐related diagnosis code. Hospitalizations for prior deliveries were eliminated. We initially used the same criteria for women with pregnancy loss outcomes (abortion or nonlive birth). However, because the lengths of these pregnancies were more varied, we then had the hospitalizations reviewed by two physicians to determine whether or not they were pregnancy related. Next, we identified prolonged hospitalizations; we assumed that lengthy hospitalizations (4 or more days) immediately before a live birth or pregnancy loss were associated with an antenatal complication. Finally, we considered hospitalizations resulting in pregnancy losses as hospitalization during pregnancy. The final categories for hospitalization during pregnancy were:
1. Live birth outcome with hospitalization occurring at least 4 days before delivery;
2. Live birth outcome with hospitalization and discharge occurring during pregnancy (women who were hospitalized and discharged while pregnant but again admitted to the hospital at least 4 days before live birth were included in this group);
3. Pregnancy loss outcome with hospitalization occurring at least 4 days before loss;
4. Pregnancy loss outcome with hospitalization and discharge occurring during pregnancy; and
5. Pregnancy loss outcome with hospitalization occurring less than 4 days before loss.
Two Institutional Review Boards (USQA Center for Health Care Research, formerly the Prudential Center for Health Care Research, and the Centers for Disease Control and Prevention) reviewed and approved the study protocol.
We included additional data from the administrative database: maternal age, delivery method, plurality, and geographic region. Delivery method was coded either vaginal or cesarean on the basis of standard diagnosis and procedure codes, plurality was coded as either singleton or multiple on the basis of International Classification of Diseases, 9th Revision codes. A unique patient identifier allowed us to examine multiple admissions for each woman in our study population.
For each antenatal hospitalization, we examined the reason for the hospitalization, the length of stay, and the associated costs. We had up to 39 diagnosis and 39 procedure codes for each hospitalization. Two obstetricians (RP, DJ) determined the primary reason for hospitalization by using a predefined list of reasons to independently categorize the diagnosis and procedure codes for each hospitalization. Any discrepancies between observations were jointly reviewed and resolved. For women who had a live birth, we subtracted any costs associated with the delivery from those of the prolonged hospitalization (occurring at least 4 days before delivery) so that the antenatal hospitalization only reflected non‐delivery‐associated charges. For women who had a pregnancy loss, we included total costs for the hospitalization (antenatal, prolonged hospitalization, and pregnancy loss).
We compared antenatal hospitalization statuses using the available covariates for the study population. Characteristics of women who were hospitalized only for a live birth (n = 42,163) were compared with women who had any hospitalization during pregnancy (including all pregnancy losses, hospitalization and discharge while pregnant, or prolonged hospitalizations). We used χ2 test to assess any significant differences between these two groups. We also examined the distribution of the number of hospital visits separately for women with live births or pregnancy losses. Finally, we examined the distribution for the cause of antenatal hospitalizations, average length of stay, and associated charges. We conducted all analyses using the Statistical Analysis Software package (SAS/STAT User's Guide, 6, 4th ed., Vol. 1, 1989, SAS Institute, Inc., Cary, NC).
Overall, 8.7% of pregnant women enrolled in a national managed care organization experienced pregnancy‐related morbidity resulting in one or more hospitalizations. Of these, 5.7% were hospitalized and discharged while pregnant, 0.8% experienced extended stays before a live birth or pregnancy loss outcome, and 2.1% experienced pregnancy loss (Table 1). The majority of women in this study were over age 30 years, had a singleton pregnancy or lived in the South. Compared with women who were only hospitalized for a live birth, women who had a hospitalization (hospitalized and discharge while pregnant, extended stay before a live birth or pregnancy loss, or pregnancy loss) were less than age 20, had a multiple birth, or lived in the northeastern United States.
Of the 4016 women who were hospitalized during pregnancy, 12.5% (472 + 29 of 4016) had more than one hospitalization during their pregnancy; the ratio of antenatal admissions was 10.1 per 100 pregnancy outcomes (4662 of 46,179) (Table 2). Among women with a live birth, 15.7% (472 of 3003) were hospitalized more than once (7.8 antenatal admissions per 100 live births, 3619 of 46,179). Of the women with a pregnancy loss outcome, 2.9% (29 of 1013) were prior to the pregnancy loss (2.3 antenatal admissions per 100 pregnancy losses, 1043 of 46,179).
Women with live birth outcomes were primarily hospitalized for preterm labor (24.4%), hyperemesis (9.3%), hypertension (9.1%), kidney disorders (6.2%), or premature rupture of membranes (6.1%) (Table 3). The overall mean length of stay for hospitalizations was 4.1 days, and the total charges were almost $28 million in 1997. Within this group, the top two reasons for delivery‐associated hospitalizations were premature rupture of membranes (35.6%) and pregnancy‐associated hypertension (27.0%) (data not shown). Among women with a pregnancy loss, 35.3% were hospitalized for ectopic pregnancy (Table 4). Pregnancy loss hospitalizations resulted in close to $9 million in charges.
The overall rate of antenatal hospitalization (10.1 per 100 deliveries) among managed care enrollees is well below the year 2000 objective of reducing hospitalizations to no more than 15 per 100 deliveries.7 It is difficult to directly compare our results with those from other studies because the populations and definitions of antenatal hospitalization have differed. However, our rate was lower than those reported in national,1,6 state,2 military,3 and clinic‐based4 studies. Our lower rate could reflect a healthier insured population that either was hospitalized at a lower rate than the general population, or it could reflect a population that used outpatient and other services (eg, home health care) available to them. Our lower rate may also be partially explained by our conservative definition of antenatal hospitalization and the nature of our inclusion criteria. Thus, it is likely that our rate of antenatal hospitalizations is an underestimate of the true maternal morbidity rate in our population. The primary causes for hospitalizations that we observed are similar to those observed in other studies.1,3,4,6,8–10 Preterm labor is one of the most common reasons for antenatal hospitalization.1,6
This is the first study that uses data from a large national managed care population to monitor inpatient care received by women during their pregnancies. (We searched the entire MEDLINE database using the terms antepartum, hospitalization, maternal morbidity, and pregnancy complications. The search was last updated when we completed final revisions to the manuscript.) Our data allowed us to examine multiple admissions, as well as pregnancy losses and live birth deliveries, in a large population of women. Moreover, because we had detailed diagnosis and procedure codes and cost data, we could further examine the reasons for and impact of antenatal hospitalizations. Despite these advantages, there were three major limitations. First, our data represented an insured population and did not include uninsured and other vulnerable populations. Second, the available information from our administrative dataset was limited. For example, we were unable to explore relationships between hospitalizations and use of outpatient care, reproductive health history, demographic factors (eg, race), and antenatal risk factors. The absence of some of this information also limits the generalizability of the findings. Moreover, we may have missed severe morbidity that resulted in immediate delivery. Finally, our sample only included women whose pregnancy ended in a hospital, regardless of outcome. Although we were able to provide some insight to pregnancy loss‐related hospitalizations, additional research is needed to accurately understand the experiences of women with a pregnancy loss who were never hospitalized.
Maternal morbidity, as measured by antenatal hospitalization, is an important health problem in terms of prevalence and costs. Further study of antenatal hospitalizations may identify specific reasons for these hospitalizations, determine whether they can be prevented, and ultimately meet our goal of decreasing maternal and fetal morbidity. Research is also needed to help determine whether hospitalizations are cost‐effective. A few studies have suggested that outpatient management of placenta previa,11–13 hypertension during pregnancy,14,15 and premature rupture of membranes16 is just as effective as inpatient management and could therefore present an alternative to hospitalization. In addition to outpatient services, expanded patient education, improved screening, and more targeted provider and patient monitoring for at‐risk groups17 may be used to possibly reduce the risk of maternal and fetal morbidity and mortality and thus the occurrence of antenatal hospitalizations.
1. Bennett TA, Kotelchuck M, Cox CE, Tucker MJ, Nadeau DA. Pregnancy-associated hospitalizations in the United States in 1991 and 1992: A comprehensive view of maternal morbidity. Am J Obstet Gynecol 1998;178:346–54.
2. Scott CL, Chavez GF, Atrash HK, Taylor DJ, Shah RS, Rowley D. Hospitalizations for severe complications of pregnancy, 1987–1992. Obstet Gynecol 1997;90:225–9.
3. Adams MM, Harlass FE, Sarno AP, Read JA, Rawlings JS. Antenatal hospitalization among enlisted servicewomen, 1987–1990. Obstet Gynecol 1994;84:35–9.
4. Haas JS, Berman S, Goldberg AB, Lee LW, Cook EF. Prenatal hospitalization and compliance with guidelines for prenatal care. Am J Public Health 1996;86:815–9.
5. Adelson PL, Child AG, Giles WB, Henderson-Smart DJ. Antenatal hospitalisations in New South Wales, 1995–96. Med J Aust 1999;170:211–5.
6. Franks AL, Kendrick JS, Olson DR, Atrash HK, Saftlas AF, Moien M. Hospitalization for pregnancy complications, United States, 1986 and 1987. Am J Obstet Gynecol 1992;166:1339–44.
7. United States Department of Health and Human Services. Healthy people 2000: National health promotion and disease prevention objectives. Washington, DC: Government Printing Office, 1990.
8. Phillippe M, Frigoletto FD, von Oeyen P, Acker D, Kitz-miller JL. High risk antenatal hospitalization. Int J Gynaecol Obstet 1982;20:475–80.
9. Acker D, Sapir J, Sachs BP, Friedman EA. Diagnostic related groups and the obstetrician: Antepartum admission. Am J Obstet Gynecol 1986;155:780–3.
10. Grimes DA. The morbidity and mortality of pregnancy: Still risky business. Am J Obstet Gynecol 1994;170:1489–94.
11. Droste S, Keil K. Expectant management of placenta pre-via: Cost-benefit analysis of outpatient treatment. Am J Obstet Gynecol 1994;170:1254–7.
12. Mouer JR. Placenta previa: Antepartum conservative management, inpatient versus outpatient. Am J Obstet Gynecol 1994;170:1683–6.
13. Love CD, Wallace EM. Pregnancies complicated by placenta praevia: What is appropriate management? Br J Obstet Gynaecol 1996;103:864–7.
14. Tuffnell DJ, Lilford RJ, Buchan PC, Prendiville VM, Tuffnell AJ, Holgate MP, et al. Randomised controlled trial of day care for hypertension in pregnancy. Lancet 1992;339:224–7.
15. Crowther CA, Bouwmeester AM, Ashurst HM. Does admission to hospital for bed rest prevent disease progression or improve fetal outcome in pregnancy complicated by non-proteinuric hypertension? Br J Obstet Gynaecol 1992;99:13–7.
16. Carlan SJ, O'Brien WF, Parsons MT, Lense JJ. Preterm premature rupture of membranes: A randomized study of home versus hospital management. Obstet Gynecol 1993; 81:61–4.
17. Brooten D, Kaye J, Poutasse SM, Nixon-Jensen A, McLean H, Brooks LM, et al. Frequency, timing, and diagnoses of antenatal hospitalizations in women with high-risk pregnancies. J Perinatol 1998;18:372–6.
© 2002 by The American College of Obstetricians and Gynecologists.