Click for next page ( 133


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 132
CHAPTER 6 The Effectiveness of Prenatal Care Inadequate or absent prenatal care is often cited for low birthweight and other poor pregnancy outcomes. prenatal care, however, cannot be determined solely on as a risk factor The value of ~ the basis of the association or ~ Is absence when Increased rise. It is possible that it is not prenatal care itself that increases the chances of a normal birthweight infant, but rather the other characteristics usually found in women who obtain such care, including optimal childbearing age, high level of education, being married, and income above the poverty line. Conversely, women who do not receive adequate prenatal care may deliver infants of low birthweight because they are characterized by other risk factors, such as extreme youth or age, poor education, being unmarried, or low income (Chapter 21. This chapter deals with the issue of whether inadequate prenatal care is an independent risk factor for low birthweight. Clearly, unless the committee could be convinced that prenatal care makes a separate contribution to birthweight, it could not claim that increasing access to prenatal care would help to reduce the incidence of low birthweight, or expect others to share its views. The committee believes that the studies reviewed in this chapter provide substantial evidence that high quality prenatal care begun early In pregnancy can lower the incidence of low birthweight. The committee reached this conclusion after looking for answers to three questions: 1. Why is it difficult to obtain valid information on the value of prenatal care in the prevention of low birthweight? 2. What factors might account for the differences in findings among studies? 3. Given the present state of knowledge, can any conclusions be reached about the value of prenatal care in the prevention of low birthweight? Issues in Studying the Effects of Prenatal Care The positive impact of prenatal care on pregnancy outcome tends to be assumed in the American health-care literature. The Select Panel 132

OCR for page 132
133 for the Promotion of Child Health concluded that prenatal care was one of the services Afar which there is such a clear consensus regarding their effectiveness and their importance to good health, that it should no longer be considered acceptable that an individual is denied access to them for any reason. . . , n ~ The Public Health Service report, Promoting Health/Prevent~ng Disease--Objectives for the Nation, was somewhat more cautious, stating that the "relative effectiveness of various interventions to improve pregnancy outcome and infant health Is not without controversy. ~2 Nevertheless, it listed as one of several objectives in the area of pregnancy and infant health an increase in the proportion of women obtaining prenatal care in the first trimester of pregnancy. Although there have been questions raised in the British literature about the effectiveness of prenatal care, particularly routine care for symptom-free women,3~5 most physicians and public health experts in the United States would agree with the sentiments expressed by the Select Panel and incorporated in the Public Health Service's objective. Researchers are aware, however, that the task of determining the effectiveness of prenatal care in the prevent ion of low birthweight, as well as other poor pregnancy outcomes, is complicated by a variety of problems. Research Des ign The f i rst problem is the ethical barr ier to conducting randomized clinical trials , which are considered the best source of evidence on the efficacy of a medical procedure. In the absence of a truly experimental model, researchers have been forced to rely on analyses of birth certificates and other large data sets and on analyses of interventions in selected populations, most frequently using comparison groups or before-after analyses. Positive findings from early studies often are difficult to interpret because significant var. tables were not controlled; because certain historical events have affected the availability of services (e.g., the passage and implementation of Medicaid); and because medical care has changed (e. g ., improvements in managing certain higher isk conditions such as diabetes and innovations in obstetr Cal practice such as electronic fetal monitor ing ~ . More thorough understanding of confounding var tables, as well as the applica- tion of more discr impinging statistical techniques, have overcome these dif f iculties to some extent. Def in ing Pr enatal Car e The Select Panel 's statement and the Public Health Service's objective both regard prenatal care as if it were a uniform entity. However, as the following review and subsequent chapters make clear, the quantity and the content of prenatal care vary widely. Although some prenatal care is clearly better than none, it also is likely that some types of prenatal care are more effective than others.

OCR for page 132
134 Research studies vary in the precision with which they def ine prenatal care. Quantitative def initions based on the number and timing of prenatal visits are more prevalent than those based on content. A few studies are limited to measur ing care versus no care, but most employ more sensitive indicators of month or tr imester when care was initiated or number of visits. These indicators may be misleading, however. Women who deliver prematurely have fewer prenatal visits than those who deliver at full term, even if they follow the recommended visit schedule until delivery; they also make a disproportionately lower number of visits because the usual schedule calls for more frequent visits later in pregnancy, by which time they have already delivered. Thus, unless a statistical ad justment is made, early - deliver ies are almost always associated with fewer prenatal visits, although the early delivery has not necessarily been caused by lack of visits. This confounding of cause and effect has been addressed in a number of ways. Ressner, most notably, developed a prenatal care index based on number of prenatal visits in relation to duration of pregnancy, the interval to the f irst visit, and type of hospital delivery service (private or general). Care was classified as adequate, intermediate, or inadequate.6 Many other investigators have used adaptations of this index, usually omitting the delivery service factor. Similarly, a report of more prenatal visits than the number recommended may be neither an indicator of excellent prenatal care nor a predictor of positive outcome. Rather, a greater-than-recommended number of visits usually indicates a high-risk pregnancy, which may have an adverse outcome despite a large number of visits. Although most current research tries to solve the quantitative issue, the problem of defining the content of prenatal care--probably of equal or greater importance--is addressed less frequently. Rle~nman stated, Rather than one specific intervention, prenatal care consists of a myr fad of interventions that are (or should be) tailored to the individual woman and her pregnancy. n7 Few studies measure those interventions, however, although some employ proxy var tables, such as Ressner 's use of delivery in a private versus a general hospital service as a measure of continuity of care. An exception is Morehead's 1970 study of health care providers.8 She developed a six-component obstetrical score for judging the quality of care based on a review of records. Although several of her components were the same quantitative ones described above, i.e., month of registration and the mumber of prenatal visits, the greatest number of points (50 out of a possible 100) was given to prenatal work-up, including history, laboratory procedures, measurements of weight and blood Pressure, nuts ition discussion, and dental care. The content of the postpartum visits, including family planning, also was scored. Research of this type is very expensive, however, because it Is labor-intensive, but may be relationship between necessary in order to obtain information about the the content of prenatal care and pregnancy outcomes.

OCR for page 132
135 Validity of Data The largest source of data on prenatal care and low birthweight consists of vital statistics reports, but the validity of the informa- tion recorded on the source documents, bir th cer tif icates, has been questioned by several researchers. In a study of North Carolina birth certif icates from 1975 through 1979, David found that birthweight was almost always recorded and was usually correct. Gestational age, however, was omitted in almost a fifth of the records and the values given were at times inconsistent with other information. Information ~ \_ ~ #a _ ~ ~ _ _ _ ~ ~ e _ _ ~ amour one earner also was missing in almost a fifth of the records.9 Whether these state findings are representative of the nation is not known. Other serious problems attend the data on the timing and quantity of prenatal care . At present, all 50 states* include month in which prenatal care began, and 49 states* include number of prenatal visits. Anecdotal reports suggest, however, that the methods by which these data are collected may be unreliable, especially for those receiving publicly funded care. For example, if a woman changes her source of care during pregnancy, only the date when she started care at the site used immediately before delivery will likely be reported, thus ignoring early care visits; and in some instances, the date of a pregnancy test is used as the date of initiation of prenatal care, which may or may not accurately reflect when actual care began. The existence of such problems is substantiated by a 1972 National Natality Survey study in which reporting of the number of prenatal visits on birth certificates was compared with survey data.~ Perfect agreement was found in only 16 percent of the cases. Land and Vaughan reviewed Missouri 1980 birth certificate data completed in hospitals using different sources of data. They found that hospitals that obtained information on prenatal care exclusively from the mother reported earlier prenatal care and more prenatal visits than those using the prenatal record only or the prenatal record and the mother. Moreover, many researchers drop from their analyses certificates on which the timing or amount of prenatal care is missing. If the number of such cases is large, this procedure may bias the results, because these women are unlikely to have the same prenatal care experiences as those for whom data are recorded, i.e., they are more likely to have little or no care; and they are unlikely to be of the same socioeconomic status. The implications of these problems are discussed by Harrison and by Showstack et ally A comprehensive study of the validity of the prenatal care items on birth certificates is essential for further high-quality research on prenatal care. Such a study should include women from a range of socioeconomic levels who receive care in a variety of settings. _, *PIUS Washington, D.C.

OCR for page 132
136 Control Variables Some of the same var tables that influence receipt of prenatal care also can influence pregnancy outcome (including complications in a prior pregnancy or low socioeconomic status); therefore, to determine the independent effects of prenatal care, it is essential to try to control for as many of these variables as possible. Most studies control for age and race at a minimum, because these clearly are associated both with receipt of prenatal care and pregnancy outcome. Other sociodemographic variables considered important include income level, education of mother and father, and mar ital status. Obstetr ical var tables also are influential, and many studies control for par ity, s ingle/multiple births, and previous pregnancy history. Although multivar late statistical measures enable more factors to be considered simultaneously than was previously feasible, it is impossible to know or include all pass ible confound ing var. tables . Select ion B ias The overr icing problem in nonrandomized studies assessing the impact of prenatal care is selection bias. The initiation of prenatal care requires action on the part of the pregnant woman. Women who take this action early usually differ from those who delay it, and in ways that can be inherently associated with r isk. Seeking prenatal care early may be a result of a woman ' s perception that she is at-r isk . For example, high-risk women may see}; prenatal care in disproportionate numbers because they may be worr fed about their pregnancies or their health, or feel ill early in their pregnancies. They may be more likely than women without anxiety or symptoms to seek early care. In many cases, the woman's anxiety may be based on family or personal exper fences that place her at h igher r isk . S. imilar ly, the symptoms that lead some women to seek car e ear ly may be caused by problems associated with low birthweight. Selection bias can also affect studies of prenatal care effectiveness when low-risk women seek early care disproportionately. Seeking prenatal care is a Good health habit." Women who seek care relatively early usually know more about health and have better general health habits than women who delay. Their knowledge and good health habits mav held to Protect their babies from low birthweight. ~ _^ , Hungary, cle ~ ay in seeking care suggests selection bias as well. For example, women who never seek care or only come for care near the end of their pregnancies may know 1 ittle about health and have poor health habits. Some women may delay initiation of care because they are dealing with one or more stressful situations, and this stress may increase their r isk of low birthweight. All studies are flawed to some extent by their inability to control for selection bias, although proxies such as maternal education may be used. Despite this bias and the other problems in studying prenatal care noted- above, policymakers must proceed on the basis of the

OCR for page 132
137 available evidence, because it is not possible to conduct randomized trials to assess the efficacy of prenatal care. Review of Studies of Prenatal Care Effects The committee reviewed two groups of studies that have attempted to determine the value of prenatal care in the prevention of low birthweight. The first group is composed of research involving large data bases, usually a year of live births for a large city, state, or county, and, in one case, information from the 1980 National Natal ity Survey (ENS). The advantage of such studies is that the large number of cases they analyze enhances the general applicability of the results and makes it possible to control for many of the variables discussed earlier and, potentially, to isolate the impact of the prenatal care. The disadvantage is the lack of precision In the measurement of prenatal care. These studies rely exclusively on data recorded on birth certificates (or the survey instruments in the NNS} for information on initiation of care and number of visits; the problems associated with these data have been noted above. Similarly, such studies provide no information on the quality or the content of the prenatal care. Women who receive inadequate care are counted in exactly the same way as those who receive high-quality care. The second group of studies analyzed by the committee includes evaluations of the impact on pregnancy outcome, and particularly on birthweight, of specific programs offering prenatal care. The advantages and disadvantages of such studies are the opposite of those described above for the large data base studies. The major disadvantage is that most of them involve a small number of births, which limits the number of var tables that can be controlled . Several other possible problems with the second group of evaluations also must be noted. Most, though not all of them, focus on low income, largely minority populations. Women in these populations are at elevated risk of delivering a low birthweight infant; but no effort has been made to determine whether programs judged ineffective for these high-risk women might be effective in reducing low b~rthweight in a lower isk population. Moreover, the positive effects reported in these studies may be influenced by the qual ity of inpatient care or other undocumented local conditions, making it difficult to generalize their findings to other settings. The advantage of these program-specific studies, however , is their greater precision in assessing the critical variable, prenatal care. A few of these studies obtain information about initiation of care and number of visits from clinic or hospital records, usually a more valid source of data than birth certificates. Most of them evaluate programs whose ob jective is to offer prenatal care appropr late to the needs of the population served. Although the content and quality of care is not usually substantiated by record reviews or other methods, there seems little doubt that most of the investigators believe they are evaluating programs offering high-quality prenatal care. Although not every women

OCR for page 132
138 in these programs may receive outstanding care, the range of quality and content is probably narrower than in the large data base studies. The results of the committee's analyses of both types of studies are summarized in the following sections. The review is limited to studies published since 1978, with the exception of the 1973 Kessne~ study,6 because this report has been cited so frequently and because its prenatal care index has been adopted by other studies. Restricting the range of articles reviewed in this way allowed the committee to focus on studies reflecting the recent improvements in prenatal care and the expansion of availability of care in the 1960s and 1970s. * More recent reports also are more likely to use stronger statistical techniques to separate the multiple effects of socioeconomic factors, medical risks, and other variables from the effects of prenatal care alone. Results of Studies Using Large Data Bases The analyses presented in Chapter 3 are the most recent in a long history of studies using local, state, and national data bases. This review begins with Kessner's study of New York City births in 1968.6 He found a s ignif icant association between adequacy of prenatal care, using var tables descr ibed earlier, and the percentage of newborns weighing 2 ,500 grams or less in each of his sociodemographic and medical-obstetric risk groups, controlling for ethnicity. Mean birthweights also varied positively with care. The strongest associations between care and birthweight were for those mothers with sociodemographic or medical r isk factors. Gortmaker reanalyzed the same data in 1979, excluding the foreign born, and reached the same conclusion using a more discriminating analytic technique that considered sociodemographic factors, gravidity, and pregnancy complications. ~ 2 He also found a significant relationship between adequacy of care and incidence of low birthweight. In addition, he noted that inadequate care was more frequently associated with low birthweight for white mothers who delivered in a genera' (nonpr ivate) service and for all black mothers. Another study of New York City was reported by Lewit in 1983 using 1970 birth certificates.~3 After controlling for sociodemographic factors, gravidity, and prior pregnancy loss, he reported a 140-gram increase in b~rthwe~ght if car e was star ted in the f ir st tr imester and a 13.6-gram increase for each visit, corrected for duration of gestation. Although no other city has been studied as intensively as New York, the data from several large counties and states have been analyzed. Showstack et al. analyzed all births in 1978 to mothers who resided in Alameda and Contra Costa counties in California.~4 Prenatal care was judged as adequate, intermediate, or inadequate on the basis of a *Even using the 1978 cut-off date, several of the studies reviewed include data from births that occurred more than 10 years ago.

OCR for page 132
139 modif ication of the Kessner or iter ia. Controlling for sociodemographic factors, multiple births, pregnancy complications, whether the birth took place in a Ra~ser-Permanente hospital, and gestational age, they found that adequate prenatal care added 197 grams to birthweight when all infants were considered. Because birthweight increased with length of gestation only until about 40 weeks, the analysis was repeated for infants of 280 or fewer days gestation. In this analysis, the association between adequate prenatal care and birthweight continued to be significant, but the impact was reduced to 100 grams for the total group and 126 grams for black infants . Harr is analyzed 1975 and 1976 data from Massachusetts, focusing particularly on blacks and taking into account sociodemographic factors, gravidity, and prior pregnancy losses He found that early initiation of care was associated with increased birthweight and that the effect was primarily through lengthening the duration of gestation. Elster examined all Utah birth certif icates from 1974 to 1979 for white women exper fencing single births who had fewer than three previous live births. ~ 6 He found that ear ly entry into prenatal care s ~ gnif icantly reduced the r isk of having a small-for-gestational age infant among pr imiparous women under age 15. Data on trends in low birthweight for four states, mentioned in Chapter 3 and documented in Appendix Table B. 13, show a decrease over the last 10 to 15 years in the rate of low-weight births regardless of the trimester in which prenatal care was initiated. In three of the states, however, the percent decline was greater for those who began care in the f irst tr imester . Many studies have used national data to examine these issues. Eisner et al. analyzed 1974 U.S. births controlling for sociodemo- graphic factors, gravidity, interpregnancy interval, and reproductive history.~7 They found that the absence of prenatal care was associated with an increased incidence of low birthweight; in fact, the authors labeled it as the greatest risk factor for low birthweight in their study. Taffel studied 1976 U.S. data controlling for educational attainment In age-race groups. She found that the incidence of low birthweight was higher in women with no prenatal care and that the proportion of low birthweight infants decreased with the frequency of visits. Greenberg analyzed 1977 U.S. birth certificates, controlling for race and maternal education, and found a significant relationship between no prenatal care and the incidence of low birthweight The absence of prenatal care had the greatest effect among black, less-educated women. Unfortunately, neither Eisner et al., Taffel, nor Greenberg controlled for gestational age in their analyses. Instead of using birth certif icate data, Rosensweig and Schultz used information from the 1967-1969 National Natality Survey (ENS) of births to married women.2Q They included the usual birth certificate variables, plus NNS information on husband's income and employment, and also added var. tables descr iptive of the health facilities and soc~oeco- nomic character istics of the study areas. They found that delay in seeking care reduced weight and gestational age at birth and that this effect was more pronounced among younger women and women of higher par ity.

OCR for page 132
140 The multivariate analysis of national data reported in Chapter 3 examined the effect of prenatal care, education, marital status, and an index of age/live birth order risk on 1981 single live births. When the timing of the first visit was used as a measure of care, elimination of late care reduced the risk of low birthweight by only 3 percent among both whites and blacks. When a modified version of the Ressner measure of the adequacy of care was used, elimination of non-adequate care reduced the low birthweight risk by 15 percent among whites and by 12 percent among blacks. The sensitivity of the results to differences in the method of measuring prenatal care suggests that the pattern of care is more significant than merely when care begins. All of these studies based on analyses of birth certificates (and in one study, data from the National Natality Survey) report one or more of these conclusions related to birthweight: . some prenatal care is better than none; early prenatal care is better than late; and adequate prenatal care is better than intermediate or inadequate care (whether adequacy is defined in terms of the time of the first visit, the number of visits in relation to gestational age, or the service setting). Finally, international data conf irm the value of prenatal care. A recent report of an international maternity care monitoring project found a positive impact of prenatal care on birthweight.2 ~ Six counts ies were studied and in four--Ch ile, Honduras, Sweden, and Thailand--there was a significant positive relationship between the number of prenatal visits and birthwe~ght after other factors were controlled. However, the study was limited to term deliveries. Results of Program Evaluations The studies reviewed in this section assess the effectiveness of special programs to provide prenatal care by comparing the pregnancy outcomes of women enrolled in the programs with the pregnancy outcomes of other women. The committee was particularly interested in programs for women at high risk because of poverty, minority status, or age, but it did not ignore programs for women generally. The ~ atter are reviewed first. The UCSF Prematurity Prevention Program One program that is of current interest is based at the University of California at San Francisco. The program includes (1) screening to detect women at high risk for preterm labor, who then receive special care; (2} education about the symptoms of preterm labor and the importance of reporting it; (3) tocolytic therapy when indicated; and (4) staff training. Herron et al. reported a reduction in the incidence of preterm deliveries at their hospital between 1977 and 1979 (the program was initiated in 1978~.22 The only comparison group was another affiliated institu-

OCR for page 132
141 tion, which did not report a similar decline in preterm deliveries. The concept of a prematurity prevention progress is being replicated In several sites using an improved research design and is described more completely in Appendix C. Health Maintenance Organizations Several studies have examined the effectiveness of care in a health maintenance organization (~O). They were intended to test the assumption that HMOS can reduce the incidence of low birthweight by encouraging early prenatal care {no financial bare iers, educational programs, etc. ~ and by providing high-quality care. Quick et al. found, however, that despite the absence of f inancial barriers, members of an HMO in Portland, Oreg., began prenatal care 1 month later and had three fewer visits than the general population of the city.2 3 This is particularly surprising because the HMO mothers tended to be older, better educated' and more often married--generally characteristics of early and frequent care seekers. For the entire cohort of 1973 and 1974 white births in the Portland area, as well as for the HMO group, the percent of infants with low birthweight increased as the level of adequate prenatal care (measured by a modification of the Ressner index) decreased. The impact of prenatal care was higher for women at elevated medical-obstetric risk. When sociodemographic and medical-obstetr ic risk factors and prenatal care were considered, HMO members had better pregnancy outcomes than did the general population. The predicted b~rthweight of infants born to HMO members was 29 grams greater than that of the general population. The relatively small size of this effect is probably due to the fact that the study population was a very healthy one. Another HMO study was conducted by Wilner et al. in Boston.24 The HMO population in this study had a higher percentage of high-risk patients (nonwhites, primigravidas, grand multiparas, and younger women) than the costar ison fee-for-service group. Enrollment for prenatal care ~ n the f irst tr imester was the same in both groups, but a significantly higher percentage of HMO patients had 11 or more visits. Although the two groups had no statistically significant differences in outcomes, including low birthweight infants, the findings suggest a positive impact of prenatal care since the EIMO group had a higher proportion of high-risk women. In the Showstack et al. study cited earl~er,~4 delivery in a Raiser-Permanente hospital was associated with a small but statistically significant increase in b~rthweight among infants of all gestations, but the difference was not significant when only the shorter gestation infants were considered. Another HMO study suggests that the content of prenatal care may be a critical factor in its potential impact on low birthweight. In a small demonstration project, Ershoff et al. reported higher mean birthweights and fewer low-weight infants among HMO mothers who received individual nutr itional counseling and a stroking cessation program, compared to a control group of TO women who did not receive such services.2 5 -

OCR for page 132
142

OCR for page 132
143 Improved Pregnancy Outcome Projects The federal government took another approach to infant health problems in 1976, when it initiated the Improved Pregnancy Outcome ~ IPO} Projects. The IPOs were to improve maternal care and pregnancy outcome In states that had contributed heavily to the incidence of infant mortality. While the federal maternal and child health agency specified many of the components of MIC projects, the states were encouraged to devise their own methods for using IPO funds. Regionalization, professional education, interorgan~zational arrangements, and other administrative approaches were emphasized, rather than direct service provision. Eventually, 34 states received IPO funds. Many difficulties were encountered in evaluating these projects.2 8 In North Carolina, IPO funds were used to develop a comprehensive care program in two counties with inadequate maternity services. Certified nurse midwives provided maternity care with assistance from local obstetr icians. The local health departments provided nutritional counseling, social services, and health education. Interdisciplinary teams planned, coordinated, and monitored patient care. Peoples et al. evaluated the effects for the period July 1, 1979, to August 30, 1981, by comparing the pregnancy outcomes of (1} all black women In the two counties served by the IPO program with those of all black women in two neighboring counties of similar socioeconomic composition; (2 ~ all black women in the IPO counties who actually registered in the IPO program with those of all black women in the comparison counties; and {3) all black teenage IPO registrants with those of all black teenagers in the comparison counties.29 On the basis of data from vital statistics, the investigators reported that the adequacy of prenatal care, as measured by an adaptation of the Kessner index, was significantly improved in all three IPO groups They did not, however , find a corresponding decrease in the incidence of low birthweight. They suggest that this may be because {1} the program did not include specific protocols for managing high-risk women (such as education of women at high-risk for preterm delivery about the early signs and symptoms of preterm labor, or interventions to decrease smoking); (2) the intensity of care was inadequate to the degree of risk; or (3) the comparison group women were at less risk and this was not completely controlled in the analysis. _ . , Community Health Centers Another federal initiative with a potential for influencing pregnancy outcome is the Community Health Center. This program evolved from the War on Poverty initiative in the mid-1960s. Today, Community Health Centers (CHCs) are private, nonprofit medical practices established by community groups receiving federal grants to provide primary health care services at reduced rates to the poor and near poor. AS of mid-1984, there were a total of 586 grantees located in all of the states and territories of the United States. Data are not available on how many of these centers provide prenatal care, nor on its content when provided, nor on how many refer pregnant women elsewhere. Of the estimated 4.5 million people currently using CHCs, however, about 29 percent are women of childbearing age (15 to 44).

OCR for page 132
144 Unfortunately, the published literature evaluating the effects of CHCs on low birthweight is scant. Two unpublished reports suggest that CHCs have an important impact on pregnancy outcome. Grossman and Goldman estimated that between 1970 and 1978, CHCS reduced the black infant mortality rate by one death per thousand live births, or 12 percent of the total decline during that per zod.3 Unfortunately, Grossman and Goldman did not collect data on low bir~chweight so it is not known whether the CHCS achieved their effect on infant mortality via birthweight improvement or other means. Schwartz and Poppen specifically analyzed the impact of CHCS on pregnancy outcomes in Baltimore in 1981.3~ Women who used Baltimore CRC s as their primary source of care were matched with women who received some care at another institutional source (excluding no care or care from a private physician). Both groups were limited to women who lived in specific census tracts, who delivered in one of seven hospitals, who were black, and who had single births. The investigators were able to show a significant effect of adequate prenatal care on birthweight and gestational age. They were unable to show any effect of receiving care at a CHC. Programs Us ing Nur se-Midwives The introduction of nur se-midwives into many service settings has provided a special opportunity to study the impact of prenatal care on under served populations. Evaluations of the impact of such service programs include those described above by Sokol et al.26 and Peoples et al.;27 a 1971 study by Levy et al. that showed a decrease in prematurity;32 a 1979 study by Reid and Morris;33 and several studies focusing on programs serving pregnant teenagers, such as that by Piechnick and Corbett.34 Most of these studies are discussed in Chapter 7. Reid and Morris compared women who delivered at Glynn-Brunswick Memorial Hospital in Georgia after the initiation of a nurse-midwife program in July 1972 with those who delivered earlier. They found a reduced incidence of low birthweight in the group served by nurse-midwives, but were uncertain whether this effect was due to the program per se. Moreover, because the nurse-midwives were providing care in an area that had a growing lack of providers generally, the study is more an anlysis of the effect on pregnancy of some care versus no care.33 OB Access The Obstetrical Access Pilot Project (OB Access) was an . attempt to improve the delivery of prenatal care to low-income' high-risk women in California. In the late 1970s, an increase in physicians refusing to accept Medi-Cal patients, coupled with increases in the number of Medi-Cal-eligible and other pregnant women reporting difficulties in obtaining prenatal care, led to the development of OB Access, funded and administered jointly by California's Medi-Cal and Maternal and Child Health programs.35 The project's goals were {1) to provide better access to comprehensive obstetrical services for Medi-Cal-elig~ble mothers in areas with inadequate obstetrical care

OCR for page 132
145 resulting from the lack of a resident obstetrician or from the decision of resident providers not to participate in Medical; and (2) to reduce per inatal mortality and morbidity rates and the percentage of preg- nancies with complications. The OB Access services included eight or more prenatal visits; nutritional and psychosocial assessments, with counseling provided to those women judged to be high risk; 16 hours of childbirth education classes; prenatal vitamins; and over 30 possible diagnostic tests, some of which (like ur ine testing) were done at most or all prenatal v' sits. By contrast, the prenatal care f inanced by Medical in the comparison group was limited to routine prenatal visits, with no reimbursement for nutr itional, educational or psychosocial services or . for prenatal vitamins. Some routine diagnostic screening was financed, but not to the extent offered in the OB Access projecte Evaluators of the OB Access Project compared project participants with a group of women whose prenatal care was reimbursed through the Medi-Cal program, matched by race/ethn~city, maternal age, par ity, plurality, sex of infant, and county of residence.36 They reported that the incidence of low birthweight was 4.7 percent for OB Access births and 7.1 percent for the Medi-Cal births. The rate of very low birthweight (less than 1,500 grams) was 61 percent lower In the OB Access group ( 1. 3 percent ver sus 0 . 5 percent}. Other Projects A variety of other demonstration projects have been or are being evaluated for their impact on pregnancy outcome, including b~rthweight. These include projects that emphasize home visiting, such as the Prenatal/Early Infancy Project in Elmira, N.Y .; a group of 10 rural infant care programs funded by the Robert Wood Johnson Foundation; and a much larger group of projects targeted specifically at pregnant teenagers, including demonstrations funded and/or evaluated by the federal Office of Adolescent Pregnancy Programs, the Ford Foundation (especially Project Redirection), the Mott Foundation through its Too Early Childbearing Network, the Ounce of Prevention Fund in Chicago, and other public and private organizations. The committee has chosen not to discuss these projects in its review of prenatal care effects for several reasons: some are ongoing programs that have not yet been evaluated; some have sample sizes that are too small; and some utilize research designs that are poorly described, weak, or constructed in a way that will not provide valid estimates of the independent impact of prenatal care on birthweight. Over the next several years, evaluations of many of the projects will be completed and will probably yield new information about how to draw women into prenatal care, the effectiveness of prenatal services generally, and how best to care for selected groups of pregnant women once they are in a prenatal system. Conclus ions This chapter reviews both the difficulties faced in assessing the value of prenatal care and two types of studies that attempt to

OCR for page 132
146 overcome these difficulties. After considering this material carefully, the committee concludes: Although a few studies have not been able to demonstrate a positive effect of prenatal care, the overwhelming weight of the evidence is that prenatal care reduces low birthweight. This finding is strong enough to support a broad, national commitment to ensur ing that all pregnant women, especially those at medical or socioeconomic r isk, receive high-quality care. * Because content of prenatal care is not def ined carefully in many of the studies reviewed, it is not possible to trace the benefits of care to specif ic aspects of the total care package. A major theme of virtually all the studies reviewed is that prenatal care is most effective in reducing the chance of low birthweight among higher isk women, whether the r isk der Eves from medical factors, sociodemographic factors, or both . Thus, dif ferences in the risk status of various study populations may partially explain var. iations in the prenatal care effects observed across studies. All of the studies reviewed that are based on large numbers of cases, particularly those using vital statistics data, show that prenatal care exerts a positive effect on birthweight. More variation exists among the results of studies evaluating special programs, although the majority show that prenatal care is associated with ~ mproved birthweight. Those special programs that have shown a positive impact on birthweight usually offer prenatal care that goes beyond more routine services to include flexible combinations of education, psycho social and nutritional services, and certain cl inical interventions such as low threshhold for hospitalization, careful screening for medical risks, and- a rapid response to the first signs of early labor. The successful projects also typically offer a package of services that is carefully defined and often described in written standards. The limited React of prenatal care Suggested by some of the special programs may result from the fact that the care was not organized to address what is now known about the causes and r isks of low birthweight. For example, the care may not have focused on such factors as smoking reduction, adequate weight gain, reducing alcohol and other substance abuse, patient and provider education about prevention of prematurity, or specific medical risks associated with low birthweight, such as bacteriuria. Unfortunately, evaluations of the smaller, more specialized programs suffer from the usual problems of studies based on quasi-experimental designs, such as self-selection and problems i n obtaining suitable comparison groups. A few of these studies also have relatively small sample sizes, which can make it more difficult to detect program effectiveness. *Steps to achieve this goal are outlined in Chapter 7, and a discussion of what high-quality prenatal care should emphasize to reduce the incidence of low birthweight appears in Chapter 8.

OCR for page 132
147 The committee believes that little will be accomplished by further efforts to document the value of prenatal care generally. Instead, more studies should be under taken to determine the effectiveness of different approaches to delivering prenatal care and of different, flexible packages of care. This issue is elaborated further in Chapter 8. References and Notes 1. Select Panel for the Promotion of Child Health: Better Health for Our Children: A National Strategy. Vol. I, p. 192. DHHS No. (PHS) 79-55071. Public Health Service. Washington, O.C.: U.S. Government Pr inting Office, 1981. Public Health Service: Promoting Health/Preventing Disease: Objectives for the Nation, p. 17. Washington, D.C.: U.S. Government Printing Office, Fall 1980. Hall ME, Chng PR, and MacGillivray I: IS routine antenatal care worthwhile? Lancet II:78-80, 1980 . 4. Oakley A: The origins and development of antenatal care. In Effectiveness and Satisfaction in Antenatal Care, edited by Enkin and I Chalmers, pp. 1-21. Philadelphia: Spastics International Medical Publications, 1982. 5. Enkin M and Chalmers I: Effectiveness and satisfaction in antenatal care. In Effectiveness and Satisfaction in Antenatal Care, edited by M Enkin and ~ Chalmers, pp. 266-290. Philadelphia: Spastics International Medical Publications, 1982. 6. Institute of Medicine: Infant Death: An Analysis by Maternal Risk and Health Care. Contrasts in Health Status, Vol. 1., edited by DM Ressner. Washington, D.C.: National Academy of Sciences, 1973. National Center for Health Statistics: Trends and variations In birthweight, p. 12. Prepared by JC Kleinman. In Health, United States, 1981. DHHS No. (PHS) 82-1232. Public Health Service. Washington, D.C.: U.S. Government Printing Office, 1981. 8. Morehead MA, Donaldson RS, and Seravalli MR: Comparisons between OEO neighborhood health centers and other health care providers of ratings of the quality of health care. Am. J. Public Health 61:1294-1306, 1971. 9. David RJ: The quality and completeness of birthweight and gestational age data in computerized birth files. Am. J. Public Health 70: 964-973, 1980. 10. National Center for Health Statistics: Comparability of Reporting Between the Birth Certificate and the National Natality Survey. Prepared by LJ Querec. Vital and Health Statistics, Series 2, No. 83. DREW No. (PBS) 80-1357. Public Health Service. Washington, D.C.: U.S. Government Printing Office, April 1980. 11. National Center for Health Statistics: Birth certificate comple- tion procedures and the accuracy of Missouri birth certificate data. Prepared by G Land and B Vaughan. In Priorities in Health Statistics: Proceedings of the 19th National Meeting of the Public Health Conference on Records and Statistics, August 1983, pp.

OCR for page 132
148 16 . 263-265. DENS No. (PHS) 81-1214. PubliC Health Service. Washington, D.C.: U.S. Government Printing Office, December 1983. 12. Gortmaker SL: The effects of prenatal care upon the health of the newborn. Am. J. Public Health 69:653-660, 1979. 13 . Lewit E : The demand for prenatal care and the production of healthy infants. In Research in Human Capital and Development, Vol. 3, edited by D Salkever, ~ Sirageldin, and A Sorkin. Greenwich, Conn.: JAI Press , 1983 . 14. Showstack JA, Budetti PP, and Minkler D: Factors associated with bir thweight: An exploration of the roles of prenatal care and length of gestation. Am. J. Public Health 74 :1003-1008, 1984 . 15. Harris JE: Prenatal medical care and infant mortality. In Economic Aspects of Health, edited by VR FUCHS, pp. 15-52. Chicago: university of Chicano Press, 1982. Elster AB: The effect of maternal age, parity, and prenatal care on postnatal outcome in adolescent mothers. Am. J. Obstet. Gynecol. 149:845-847, 1984. 17. Eisner V, Brazie JV, Pratt MW, and Hexter AC: The risk of low birthweight. Am. J. Public Health 69:887-893, 1979. 18. National Center for Health Statistics: Prenatal Care: United States, 1969-1975 . Prepared by S Taffel. Vital and Health Statistics, Series 21, No. 33. DHE:W No. (PENS) 78-1911. Public Health Service. Washington, D.C.: U.S. Government Printing Off ice, September 1978 . 19. Greenberg RS: The impact of prenatal care in different social groups. Am. J. Obstet. Gynecol . 145: 797-801, 1983 . 20. Rosenzweig MR and Schultz TP: The behavior of mothers as inputs to child health: The determinants of birth weight, gestation, and rate of fetal growth. In Economic Aspects of Health, edited by VR Fuchs, pp. 53-92. Chicago: University of Chicago Press, 1982. Donaldson W and Billy JOG: The impact of prenatal care on birth weight: Evidence from an international data set. Med. Care 22:177-188, 1984. Herron MA, Katz M, and Creasy RK: Evaluation of a preterm birth prevention program: Preliminary report. Obstet. Gynecol. 59:452-4S6, 1982. Quick JO, Greenlick MR, and Roghmann KJ: Prenatal care and pregnancy outcome in an HMO and general population: A multivar late cohort analysis. Am. J . Public Health, 71: 381-390, 1981. 24. Wilner S. Schoenbaum SC, Monson RR, and Winickoff RN: A comparison of the quality of maternity care between a health maintenance organization and fee-for-serv~ce practices. N. Engl. J. Med. 304:784-787, 1981. Ershoff OH, Aaronson NO, Danaher BG, and Wasserman FW: Behavioral, health and cost outcomes of an MO-based prenatal health education program. Public Health Rep. 98:536-547, 1983. Sokol RJ, Woolf RB, Rosen MG, and Weingarden R: Risk, antepartum care, and outcome: I - act of a maternity and infant care project. Obstet. Gynecol. 56 :150-156, 1980.

OCR for page 132
149 27 . Peoples MD and Siegel E: Measur ing the impact of programs for mothers and infants on prenatal care and low birthweight: The value of refined analyses. Med. Care 21: 586-605, 1983. 28. Strobino DM: Is it Possible to Evaluate the IPO Project? Am. J. Public Health 74:541-542, 1984. 29. Peoples MD, Grimson RC, and Daughty GL: Evaluation of the effects of the North Carolina improved pregnancy outcome project: I - locations for state-level decision-making. Am. J. Public Health 74: 549-554, 1984. 30. Grossman M and Goldman F: An Economic Analysis of Community Health Centers: Final Report. New York: National Bureau of Economic Research, 1982. 31. Schwartz R and Poppen P: Measuring the Impact of CHCS on Pregnancy outcomes: Final Report. Cambridge, Mass.: ABT Associates, 1982. 32. Levy BS, Wilkinson FS, and Marine WH: Reducing neonatal mortality rate with nurse-midwives . Am. J. Obstet. Gynecol. 109:50-58, 1971 33. Reid ML and Morris JB: Prenatal care and cost-effectiveness: Changes in health expenditures and bir th outcome following the establishment of a nurse-midwife program. Med. Care 17:491-500, 1979. 34. Piechnik SL and Corbett MA: Adolescent pregnancy outcome: An experience with intervention. J. Nurse-Midwifery, in press. 35. Maternal and Child Health Branch: Final Evaluation of the Obstetrical Access Pilot Project, July 1979 to June 1982. Sacramento, Cal~f.: Department of Health Services, 1984. Korenbrot CC: Risk reduction in pregnancies of low-income women: Comprehensive prenatal care through the OB Access Project. Mobius 4: 34-43, 1984 . .