Prenatal Care and Length of Gestation - NCBI

1 downloads 0 Views 1MB Size Report
certificate data were studied for babies born in 1978 to mothers who were residents of Alameda or Contra Costa counties, California. Using multiple regression ...
Factors Associated with Birthweight: An Exploration of the Roles of Prenatal Care and Length of Gestation JONATHAN A. SHOWSTACK, MPH, PETER P. BUDETTI, MD, JD,

AND

DONALD MINKLER, MD

Abstract: To assess the association with birthweight of prenatal medical care, length of gestation, and other prenatal factors, birth certificate data were studied for babies born in 1978 to mothers who were residents of Alameda or Contra Costa counties, California. Using multiple regression data analytic techniques, adequate prenatal care (defined by the number of prenatal care visits compared to length of gestation and month of start of care) was found to be associated with an increase of 197 grams in average birthweight. This effect was even greater for Black infants and infants of short

length of gestation. Adding length of gestation to the equation increased significantly the proportion of the variance in birthweight accounted for. For babies of short gestation (s 280 days), the addition of length of gestation was associated with a halving of the association of prenatal care with birthweight. The results suggest that researchers need to take into account the nonlinear relationship between length of gestation and birthweight when assessing factors that affect birthweight. (Am J Public Health 1984; 74:1003-1008.)

Introduction There is a strong statistical relationship between higher birthweight and better newborn survival. ' In recent years, a number of studies have found a positive relation between prenatal care and birthweight. In 1966, Abramowicz and Kass concluded that the published evidence supporting a positive relation between prenatal care and birthweight was inconclusive.2 Using data from the 1972 National Natality Survey, Placek found that absence of prenatal care was associated with low birthweight.3 Studying birth certificate data, Taffel found that a higher number of prenatal visits was associated with normal birthweight, but that mother's education and ethnic group were more important than the timing of the onset of prenatal care in determining low birthweight.4,5 By studying 130,000 births in upstate New York in 1973, Stickle and Ma found that prenatal care that started in the first trimester was associated with better pregnancy outcome.6 Gortmaker studied vital statistics records for all births in New York City in 1968.7 He found that-after controlling for a variety of factors, including parent's education, maternal age, length of gestation and adequacy of prenatal care based on Institute of Medicine criteria8-adequate prenatal care was associated with higher birthweights. He also found that only 45.8 per cent of White infants and 10.7 per cent of Black infants had "adequate" prenatal care. Quick and colleagues found that, in 1973-74 in Portland, Oregon, both HMO (health maintenance organization) membership and prenatal care (using the Institute of Medicine definition) were positively associated with birthweight.9 Sokol, et al, reported significantly better perinatal outcomes, including higher birthweight, for women who received comprehensive prenatal care than for women of

similar risk who received standard care. 10 Greenberg demonstrated an association between prenatal care and higher birthweight that varied somewhat but persisted after controlling for mother's race and education." Peoples and Siegel reported a significant association between adequacy of care and the proportion of infants less than or equal to 2500 grams birthweight when controlling for source of care, reproductive risk, and marital status.'2 This paper confirms and extends these earlier findings with particular emphasis on an exploration of the relationship between prenatal care, length of gestation, ethnicity, and birthweight. A major methodological difference that distinguishes this work is that nonlinearities in the relation between length of gestation and birthweight are taken into account when testing for the association of prenatal care with birthweight.

From the Institute for Health Policy Studies, and the Departments of Pediatrics, and of Obstectrics, Gynecology and Reproductive Sciences, School of Medicine, University of California, San Francisco. Address reprint requests to Jonathan A. Showstack, MPH, Assistant Adjunct Professor of Health Policy, Institute for Health Policy Studies, University of California, 1326 Third Avenue, San Francisco, CA 94143. Dr. Budetti is Associate Professor of Social Medicine in Pediatrics; Dr. Minkler is Professor of Obstetrics, Gynecology and Reproductive Sciences. This paper, submitted to the Journal July 22, 1983, was revised and accepted for publication March 1, 1984.

©1984 American Journal of Public Health 0090-0036/84 $1.50

AJPH September 1984, Vol. 74, No. 9

Methods Data Source

Data were obtained from California's Maternal and Child Health Data Base, derived from vital statistics records.'3 All births during 1978 to mothers who were residents of Alameda or Contra Costa counties were studied. Data for 1978 were chosen because only during that year did California birth records include additional crucial data, items such as the number of prenatal visits. There were 25,091 births to mothers who were residents of Alameda or Contra Costa counties during 1978. Edits of the files to exclude cases with outlying values or missing data on key variables reduced the number of cases for most analyses to 18,470 or fewer. For example, cases were omitted from our analyses if data were missing on birthweight or if birthweight was listed on the birth certificate as less than 500 grams or more than 6,000 grams. Compared to birth certificates from other hospitals, birth certificates from Kaiser-Permanente hospitals were more likely to include complete information on the variables studied (in particular, length of gestation) and were thus more likely to be included in our analysis.* Thus, the study population differed slightly from mothers in these two counties with maternal age being slightly higher and with fewer Black and other minority mothers. Because there is a relatively good distribution of *89 per cent inclusion for Kaiser-Permanente hospital birth certificates compared to 76 per cent inclusion for other birth certificates.

1 003

SHOWSTACK, ET AL.

demographic characteristics across Kaiser and non-Kaiser hospitals, this differential exclusion should not have affected our results substantially. It is quite possible, however, that cases excluded because of missing data** on length of gestation may have been a higher risk population.'4 See below for a discussion of this issue. Data recorded on birth certificates are not likely to be as accurate as data collected directly from medical records. Random errors in data will tend to weaken the results, while non-random errors will tend to bias the results. Studies of birth certificates imply that the error rates are acceptable for this type of study. Of most importance, birthweight and trimester of the start of prenatal care are generally reported relatively accurately.'5 A systematic bias in the recording of any of the variables studied seems unlikely, but since it remains a possibility our results should be interpreted with due caution. The definitions of most of the variables studied are fairly straightforward, with one exception. "Complications of pregnancy" are physician-defined and, thus, are based on a physician's own judgment. Our primary results remain essentially the same whether this variable is included or excluded from the analyses. Data Analysis

It was hypothesized that each of a set of variables obtained from birth records would be associated independently with birthweight. The variables studied were the

infant's sex and ethnic group, whether the infant was the

product of a multiple birth, the mother's age, and education, whether there were complications of the pregnancy (as defined by the physician in charge), adequacy of prenatal care, type of hospital of birth (Kaiser-Permanente Health Plan or other), and length of gestation. The adequacy of prenatal care was determined according to a slight modification of the definition developed by the Institute of Medicine8 (see Appendix). Prenatal care is characterized as "adequate," "intermediate," or "inadequate" depending on the number and timing of prenatal visits. For example, for prenatal care to be adequate, it must have been initiated in the first trimester and have had a sufficient number of visits for a specific length of gestation. The quality of the prenatal care received is not measured. In contrast to the definition of adequacy of prenatal care developed by the Institute of Medicine, in which cases that had missing data on number of prenatal visits were assigned to the "inadequate" prenatal care group, we excluded cases that had missing data on any of the variables from which adequacy of prenatal care was computed. An inspection of the raw data showed that most cases excluded from our analyses had missing data on the length of gestation and, based on their values for the month that prenatal care started and the number of prenatal care visits, would probably have been assigned to "adequate" prenatal care had data for length of gestation been present. Because only approximately 3 per cent of the cases had data missing on number of prenatal visits, had these cases been assigned to "inadequate" prenatal care according to the Institute of Medicine definition rather than omitted from our analysis, our results "*Note that the number of subjects shown in the Tables varies because subjects with missing data on a particular variable displayed in a Table are omitted from that Table. The data reported in Tables 1, 4, and 5 come from the most restricted data sets, used in the multiple regression analyses, where missing data on any one of the many variables studied would cause the case to be dropped from that analysis.

1004

on per cent of the population receiving adequate prenatal care would have changed only slightly. Although it would have been possible to identify public hospital births, births in a public hospital were relatively few (approximately 5 per cent of the population) and were, in general, limited to women of lower socioeconomic status. Because of this correlation with other socioeconomic variables in the analysis, it was decided to study another hospital characteristic of significance in this geographic area, whether the hospital belonged to the Kaiser-Permanente health maintenance organization or not. The principal analytic techniques were chi-square tests for categorical data and linear multiple regression analysis. Regression analysis was used to study the independent association between birthweight and various prenatal factors. The independent variables studied by regression analysis are listed in Table 1. An inspection of the first-order correlation matrix showed that collinearity was generally not a problem for the analysis. "Adequate" and "intermediate" prenatal care, however, were highly correlated with each other (r = .90), while hospital type (Kaiser versus non-Kaiser) had a firstorder correlation with the prenatal care variables that, although small, would tend to cause collinearity problems (e.g., hospital type had a first-order correlation with both birthweight and adequate prenatal care of .07). The potential effect of these correlations is to weaken the test of the relation between birthweight and both prenatal care and hospital type. While these correlations would not affect the estimates of the degree of association of, for example, "adequate" and "intermediate" prenatal care with birthweight, they would tend to inflate the error terms for these estimates, thus reducing the power of the test that these two variables were each significantly different from the reference group, "inadequate" care. Thus, the statistical significance of the relationship of birthweight to adequacy of prenatal care and hospital type may be somewhat understated in the results described below. Ethnicity was coded (according to birth certificate definitions) into four mutually exclusive groups (White, Black, Mexican-American, and all others). Note that when more than two groups are coded, one group is omitted in the regression equations and serves as a reference; thus, TABLE 1-Characteristics of Births Studied (n = 18,243)

Birthweight (mean grams)

3363 grams

Sex (male) Multiple Birth Ethnic Group White Black Mexican-American Other Mother's Age and Education 17 years old 18 to 34 years old, education - 1 1 years 18 to 34 years old, education .12 years .35 years old, education - 11 years .35 years old, education .12 years Complications of Pregnancy Hospital Type (Kaiser) Prenatal Care Adequate Intermediate Inadequate Gestational Age (mean days)

51 2 61 20 10 9

4 12 79 1 5 8 37 72 24 4 283 days

AJPH September 1984, Vol. 74, No. 9

FACTORS ASSOCIATED WITH BIRTHWEIGHT

"White" ethnicity is the reference group to which the other three ethnic groups are compared. To assess the interaction of mother's age (less than 18 years old, 18 to 34, 35 or above) and education (high school graduate or not), five mutually exclusive age-education groups were coded. Mothers 17 years old or younger were assumed not to be high school graduates. The age-education reference group was mothers between 18 and 34 years old who were high school graduates.

Results The characteristics of the births studied are shown in Table 1. Of note are the high proportion of births to mothers who had adequate prenatal care (72 per cent) and of births in hosptials of the Kaiser-Permanente health plan (37 per cent). The rate of adequate prenatal care was high for all ethnic groups, with Whites, Mexican-Americans, Others, and Blacks receiving adequate care in that order (Table 2). Babies with adequate prenatal care had an (unadjusted) mean birthweight over 200 grams higher than babies who had intermediate or inadequate prenatal care (3419 grams versus 3206 grams). This effect was most striking in Black babies. In contrast to the experience reported by Quick, et al,9 from the Kaiser health plan in Portland, Oregon, babies born in Kaiser hospitals in the area that we studied were significantly more likely to have had adequate prenatal care than were babies born in non-Kaiser hospitals (Table 3). To assess the independent association between birthweight and each of the characteristics of the births being studied, several multiple regression equations were computed. The results of these analyses show that, holding other variables constant statistically, the following variables demonstrated significant independent associations with birthweight: infant's sex, multiple birth, Black ethnicity, maternal age and education, complications of pregnancy, birth in a Kaiser hospital, gravidity, adequate prenatal care, and length of gestation. Table 4 shows the association of these characteristics with birthweight when length of gestation is and is not part of the equation. Length of gestation accounts for an additional 4 per cent of the variance in birthweight and adds, on average, 3.7 grams to birthweight per day of gestation. Of note in this study, as well as in other similar analyses of the linear relationship between birthweight and various TABLE 2-Mean Birthweight According to Adequacy* of Prenatal Care and Ethnicity of Baby

Prenatal Care" Intermediate or Inadequate

Adequate

Mean

Mean

Ethnicity

Birthweight

N

Birthweight

N

White Black MexicanAmerican Other TOTAL

3481 3226

8658 (75)"** 2323 (61)

3294 3005

2850 (25) 1468 (39)

3295 3184 3206

602 (34) 566 (34) 5486 (29)

3435 3331 3419

1182 (66) 1098 (66) 13261 (71)

*See text for a description of the definition of adequacy of prenatal care. **Chi-square = 311.83, df = 3, p < .001. ***

Parentheses enclose row percentages.

AJPH September 1984, Vol. 74, No. 9

TABLE 3-Adequacy of Prenatal Care by Hospital Type Prenatal Care* Hospital Type

Adequate

Intermediate

Inadequate

N

Kaiser Non-Kaiser

75.4 68.0

21.3 25.4

3.3 6.6

11,881

6,866

*See text for a description of the definition of adequacy of prenatal care. Chi-square = 153.14, d.f. = 2, p < .001.

independent variables, is the relatively low per cent of the variance in birthweight accounted for by the model studied. The relation between birthweight and length of gestation is not linear, however. Birthweight rises on average with length of gestation only until the gestation reaches about 40 weeks, after which the relationship levels off and even becomes slightly negative for longer gestations. 16 Therefore, one might postulate that any effect on birthweight of length of gestation will be apparent, for the most part, only in babies with gestational ages of 40 weeks or less. Likewise, if prenatal care affects birthweight by lengthening gestation, this effect should be most apparent in babies with shorter gestations (that is, the effect would tend to "wash-out" statistically if babies with longer gestations are included in the analysis). There are several ways that this might be modeled statistically. One might analyze only births of 40 weeks or less gestation, or one might truncate the gestational length distribution so that, for example, all births of greater than 40 weeks gestation are given a value in the analysis of 40 weeks gestation. The latter method assumes that the average effect of gestational age on birthweight lessens in importance after 40 weeks gestation compared with the effect on birthweight of other variables. We analyzed our data in both these ways and display the results for the short length of gestation group (equal to or less than 280 days gestation) in Table 5. Our hypothesis about the birthweight/gestational age relationship is supported by the analysis. For babies with short lengths of gestation, we are able to account for 34 per cent of the total variance in birthweight compared to 18 per cent for the total gestation group. For Black babies the effect is even more pronounced, with 41 per cent of the variance accounted for. For the short gestation babies, each day of gestation is associated with an increase in birthweight of 14 grams. The data in Tables 4 and 5 suggest that the association between birthweight and prenatal care remains even after taking length of gestation into account, but that this association is substantially less with length of gestation in the equation. Thus, in Table 4 when length of gestation is added to the equation the impact of adequate prenatal care drops 10 grams (207 grams to 197 grams). When we turn to Table 5, however, where we examine babies with shorter lengths of gestation, the association of adequate prenatal care with birthweight drops 113 grams (223 grams to 100 grams) once length of gestation is accounted for. This effect on the strength of the association between birthweight and prenatal care once gestational age is added to the equation is intriguing and may be explained in several different ways. Two plausible but somewhat contradictory explanations are: 1) that it reflects a poorly specified model when length of gestation is not in the equation (as suggested 1 005

SHOWSTACK, ET AL. TABLE 4-The Association of Characteristics of Birth with Birthweight (in grams) for Babies Born to Mothers Who Were Residents of Alameda or Contra Costa counties, California, 1978 Without Length of Gestation as an Independent Variable n= 18,243

Independent Variable

Sex (male = 1) Multiple Birth (yes = 1) Ethnic Group Black Mexican-American Other Mother's Age and Education s17 years old 18 to 34 years, education ± 1 1 years .35 years, education ±11 years .35 years, education .12 years Complications of Pregnancy (yes = 1) Hospital Type (Kaiser = 1) Prenatal Care Adequate Intermediate Length of Gestation (days) Constant Adjusted R2

B

S.E.

With Length of Gestation as an Independent Variable n= 18,243 B

S. E.

+130 -921

8.26*** 29.25***

+132 -858

8.06*** 28.59***

-240 -20 -149

10.75*** 14.67 14.83***

-228 -21 -133

10.49*** 14.30 14.47***

-79 -40 +42 +39 -282 +57

21.91** 13.41 47.56 20.00 15.46*** 8.67***

-71 -39 +52 +42 -257 +52

21.36*** 13.07** 46.38 19.50* 15.09*** 8.45***

+207 +65

21.06*** 21.90**

+197 +71 +3.7 2185 .18

20.53*** 21.35*** 0.12***

3222 .14

*p < .05.

"p < .01. -p < .oo1. NOTE: All variables, except birthweight and length of gestation, are in the form of 0,1 dummy variables. For dichotomous variables, the value for "1" is given in the Table and the value for "0" can be inferred. For Ethnic Group, the reference group was "White," for the

mother's age and education, the reference group was "age 18-34, education .12 years," and for prenatal care, the reference group was "inadequate." For example, holding all other variables constant statistically, boys weighed on average 132 grams more than girls at birth when gestational length is accounted for.

by the model's low R2 value); or 2) that it occurs because the effect of prenatal care is partially exerted through length of gestation itself (that is, prenatal care lengthens gestation, which in turn raises birthweight).*** This association between adequate prenatal care and longer gestation is suggestive, but does not prove a causal relationship. Our second method of analyzing the nonlinear relation between birthweight and length of gestation described above (by truncating the entire distribution) showed results that were similar to, but slightly less pronounced than, the first method. For example, the amount of variance accounted for in the second method is 30 per cent compared to 34 per cent by the first method. Since the baby's ethnic group and the amount of prenatal care received tended to be correlated, we were unable to test directly whether adequate prenatal care produced different results for different ethnic groups. We did, however, compute separate regression equations for each ethnic group. For the shorter gestation group, adequate prenatal care was associated with an increase in birthweight of 126 grams for Black babies and 105 grams for White babies. ***This latter inference assumes that prenatal care may affect length of gestation and not vice versa. By comparing the number and timing of prenatal visits to length of gestation, our definition of the adequacy of prenatal care should allow this type of inference, that is, there should not be a spurious relationship due to the number of visits being a proxy for length of gestation or vice versa. Controlling for gestational age at birth allows comparison of the association between birthweight and different patterns of prenatal care for infants at the same level of development. Since most babies who were classified as receiving inadequate prenatal care had either no care with normal gestation or started care in the third trimester and had very few visits, it is unlikely that "inadequate" prenatal care is a proxy for short gestation. 1006

To test the effects on birthweight of different definitions of adequate prenatal care, we also computed a regression equation for the entire cohort of births defining prenatal care as adequate if it had started in the first trimester. This equation showed an association between adequate prenatal care and birthweight of only 35 grams on average, 172 grams less than the effect on birthweight shown by the definition that compared prenatal visits with gestational age to define the adequacy of prenatal care. Discussion Results of this study provide additional evidence that prenatal care, especially after taking into consideration the number and timing of prenatal visits, has a substantial association with higher birthweight, especially for Black babies. This association is, on average, approximately 6 per cent of birthweight for the total group and 8 per cent of birthweight for Black babies and persists when controlling for a variety of other independent variables, including length of gestation, sex, multiplicity, mother's age, education and gravidity, complications of pregnancy, and type of hospital of birth. Our results also suggest that investigators need to take into account the nonlinear relation between length of gestation and birthweight when assessing factors that are associated with birthweight. By taking into account nonlinearities in the relation between length of gestation and birthweight, we were able to account for a substantially higher proportion of the variance in birthweight than in other studies, and we were able to estimate an increment in average birthweight that may have AJPH September 1984, Vol. 74, No. 9

FACTORS ASSOCIATED WITH BIRTHWEIGHT TABLE 5-The Association of Characteristics of Birth with Blrthwelght (in grams) for Babies Bom to Mothers Who Were Residents of Alameda or Contra Costa counties, California, 1978: Babies of 280 or fewer Days Gestation only

With Length of Gestation as an Independent Variable

Without Length of Gestation as an Independent Variable n = 8455 Total Group Independent Variable

Sex (male = 1) Multiple Birth (yes = 1) Ethnicity Black Mexican-American Other Mother's Age and Education s17 years old 18 to 34 years, education 11 years -35 years, education s1 1 years .35 years, education >12 years Complications of Pregnancy (yes = 1) Hospital Type (Kaiser = 1) Prenatal Care Adequate Intermediate Length of Gestation (days) Constant Adjusted R2

S.E.

B

n = 8455 Total Group B

n = 1999 Black Infants Only

S.E.

B

-815

12.65"* 34.92***

+125 -619

11.28*** 31.41*

-208 +3 -89

15.69*** 22.80 21.10***

-164 +8 -89

14.02*** 20.32 18.81*

-62 -4 +96 +47 -399 +22

32.49 20.14 66.44 28.66 21.46*** 13.47

-42 +6 +126 +52 -229 +21

28.96 17.95 59.22* 25.55* 19.47*** 12.00

-46 +39 +351 +3 -243 +51

+223 +42

30.84* 31.83

+100 +25 +14 -735

27.61* 28.37 0.31**

+126 +59 +17 -1670

+121

3051 .17

S.E.

+99 -473

.34

22.97*** 59.89***

41.91 37.29 163.94* 67.39 36.1 1* 25.91

47.76** 48.92 0.58*** .41

.p < .05.

^^p < .01.

^^^p < .oo1. NOTE: All variables except birthweight and length of gestation are in the form of 0,1 dummy variables. For dichotomous variables, the value for "1" is given in the Table and the value for "0" can be inferred. For Ethnic Group, the reference group was "White," for the mother's age and education, the reference group was "age 18-34, education .12 years," and for prenatal care, the reference group was "inadequate."

been due to the association of prenatal care with gestational age. Our data also suggest that failure of some earlier studies to find a substantial relation between prenatal care and birthweight may have been due to not taking into account the number and timing of prenatal care visits. When our data were analyzed using a simple definition of adequate prenatal care as being care that started in the first trimester, the association with birthweight was reduced substantially. This illustrates the importance for public policy debate of collecting and analyzing these data elements routinely on birth certificates. In recent years, for example, California first added, then dropped, key data elements from its vital records, thereby increasing, and then decreasing, the usefulness of these records for research purposes. This analysis was based on a relatively sophisticated definition of adequate prenatal care. We were not able, however, to determine the content or quality of the prenatal care delivered or some characteristics of the women whose prenatal care was "adequate." In addition, although we attempted to control for other factors that may have been associated with both adequate prenatal care and birthweight, this was not a study of women assigned randomly to prenatal care. It is possible that the positive results for prenatal care may have been spurious due to lack of measurement of an intermediate variable that was highly associated with adequate prenatal care, length of gestation and birthweight. Although imperfect, our adjustment for ethnic group, mother's age and education, and membership in a health maintenance organization likely captured much of the variance that would have been due to socioeconomic status had it been measured directly. We were unable to measure either directAJPH September 1984, Vol. 74, No. 9

ly or indirectly, however, such factors as whether the pregnancy was planned, the mother's smoking habits, substance abuse, stress, or lifestyle. While analysis of vital statistics can point us in the right direction concerning the factors that affect perinatal outcome, there is a need for studies in greater depth. Of particular importance are studies that identify the specific aspects of prenatal care that may lead to improved birthweight and thus improved perinatal outcome. In particular, if prenatal care lengthens gestation, is this effect confined to a few women who would have delivered their babies extremely early or does it include a mechanism that lengthens gestation somewhat for all babies? Although we may be able to identify proxy factors associated with higher birthweight, this study points out the general lack of understanding of the biologic, social, and medical care mechanisms that such proxies represent.

REFERENCES 1. Williams RL, Chen PM: Identifying the sources of the recent decline in perinatal mortality rates in California. N Engi J Med 1982; 306:207-214. 2. Abramowicz M, Kass EH: Pathogenesis and prognosis of prematurity. N Engl J Med 1966; 275:878-885, 938-943, 1001-1007, 1053-1059. 3. Placek PH: Maternal and infant health factors associated with low infant birthweight: findings from the 1972 National Natality Survey. In: Reed DM, Stanley FJ, (eds): The Epidemiology of Prematurity. Baltimore: Urban & Schwarzenberg, 1977. 4. Taffel S: Prenatal Care United States, 1969-1975. Vital and Health Statistics, Series 21, No. 33, DHEW Pub. No. PHS 78-1911. Hyattsville, MD: National Center for Health Statistics, 1978. 5. Taffel S: Factors associated with low birthweight United States, 1976. Vital and Health Statistics, Series 21, No. 37, DHEW Pub. No. (PHS) 801915. Hyattsville, MD: National Center for Health Statistics, 1980.

1007

SHOWSTACK, ET AL. 6. Stickle G, Ma P: Some social and medical correlates of pregnancy outcome. Am J Obstet Gynecol 1977; 127:162-166. 7. Gortmaker SL: The effects of prenatal care upon the health of the newborn. Am J Public Health 1979; 69:653-660. 8. Kessner DM, Singer J, Kalk CE, Schlesinger ER: Infant death: an analysis by maternal risk and health care. In: Contrasts in Health Status, Vol I. Washington, DC: Institute of Medicine, National Academy of Sciences, 1973. 9. Quick JD, Greenlick MR, Roghmann KJ: Prenatal care and pregnancy outcome in an HMO and general population: a multivariate cohort analysis. Am J Public Health 1981; 71:381-390. 10. Sokol RJ, Woolf RB, Rosen MG, Weingarden K: Risk, antepartum care, and outcome: impact of a maternity and infant care project. Obstet Gynecol 1980; 56:150-156. 11. Greenberg RS: The impact of prenatal care in different social groups. Am J Obstet Gynecol 1983; 145:797-801. 12. Peoples MD, Siegel E: Measuring the impact of programs for mothers and infants on prenatal care and low birth weight: the value of refined analyses. Med Care 1983; 21:586-605. 13. Williams RL: Measuring the effectiveness of perinatal medical care. Med Care 1979; 17:95-1 10. 14. David RJ: The quality and completeness of birthweight and gestational age data in computerized birth files. Am J Public Health 1980; 70:964-973. 15. Jamison H: Birth certificate data quality: an exploratory study. Sacramento: Center for Health Statistics, California Department of Health Services, 1983. 16. Hoffman H, Lundin FE, Bakketeig LS, et al: Classification of births by weight and gestational age for future studies of prematurity. In: Reed DW, Stanley FJ (eds): The Epidemiology of Prematurity. Baltimore: Urban and Schwarzenberg, 1977.

San Francisco for their comments on an earlier draft of this paper. This research was supported in part by a grant from The National Center for Health Services Research (No. HS02975).

APPENDIX Definition of Adequacy of Prenatal Care*

Definition Adequate (care initiated in the first trimester and)

ACKNOWLEDGMENTS

I

Then Number of Prenatal Visits Must Be:

s13 14 to 17 18 to 21 22to25 26to29 30 to 31 32 to 33 34to35

.36 Inadequate (care initiated in the third trimester or) Intermediate

The authors would like to acknowledge Carol Korenbrot, PhD, and other colleagues at the Institute for Health Policy Studies, University of California,

If Gestation is: (weeks)

.1 .2 .3 .4 .5 .6 .7

.8 .9

14 to 21 0 22 to 29 s1 30 to 31 s2 32to33 -3 -34 s4 All combinations other than above.

'Based on Institute of Medicine definition8; however, cases with missing data on gestational length, number of prenatal visits, or trimester of initiation of prenatal care are excluded.

Seminars Announced on Mine Safety and Health Law

The Energy Bureau has announced a new series of seminars on federal mine safety and health law. Entitled "Mine Safety And Health Law: New Developments, New Defenses," these programs will be held on October 1st at the New Orleans Marriott Hotel and on October 22nd at the Denver Marriott Hotel-City Center. The program, created for officials or attorneys involved in any way in mine safety and health laws and regulations, will bring attendees up to date on all aspects of federal mine safety and health law and deal with the many changes that have occurred since the most recent major legislation, in 1976. Seminars will cover such areas as: * the new administration at the Mine Safety and Health Administration, its new leader and his staff; * the rapidly growing area of discrimination and miners' rights; * mandatory standards; * enforcement developments and defenses; * the Federal Mine Safety and Health Review Commission, its jurisdiction, practices and procedures; * civil and criminal liability of supervisors; and * all other aspects of mine safety and health law and the developments and defenses arising from its changing face. For further information, contact Robert W. Nash, president, The Energy Bureau, Inc., 41 East 42nd Street, New York, NY 10017. 212-687-3177.

1008

AJPH September 1984, Vol. 74, No. 9