Abstract
We examined the relation between maternal smoking and adverse infant outcomes [low birth weight (LBW), and preterm birth (PTB)] during 2007–2008 in San Bernardino County, California—the largest county in the contiguous United States which has one of the highest rates of infant mortality in California. Using birth certificate data, we identified 1,430 mothers in 2007 and 1,355 in 2008 who smoked during pregnancy. We assessed the effect of never smoking and smoking cessation during pregnancy relative to smoking during pregnancy for the 1,843/1,798 LBW, and 3,480/3,238 PTB’s recorded for 2007/2008, respectively. To describe the effect of quitting smoking during pregnancy, we calculated the exposure impact number for smoking during pregnancy. Major findings are: (1) relative to smoking during pregnancy, significantly lower risk of LBW among never smoking mothers [OR, year: 0.56, 2007; 0.54, 2008] and for smoking cessation during pregnancy [0.57, 2007; 0.72, 2008]; (2) relative to smoking during pregnancy, significantly lower risk of PTB was found for never smoking mothers [0.68, 2007; 0.68, 2008] and for smoking cessation during pregnancy [0.69, 2007; 0.69, 2008]; (3) an exposure impact assessment indicating each LBW or PTB outcome in the county could have been prevented either by at least 35 mothers quitting smoking during pregnancy or by 25 mothers being never smokers during pre-pregnancy. Our findings identify an important burden of adverse infant outcomes due to maternal smoking in San Bernardino County that can be effectively decreased by maternal smoking cessation.
Keywords: San Bernardino County, Maternal tobacco use, Smoking cessation during pregnancy, Exposure impact number
Introduction
Among women in the United States, maternal cigarette smoking is one of the most important modifiable risk factors for adverse infant outcomes (e.g., low birth weight (LBW), preterm-birth (preterm), and sudden infant death syndrome (SIDS)) [1–3]. Although smoking rates among women have been decreasing in the United States, an estimated 22 % of women of reproductive age continued to smoke in 2006 [4], and it is estimated that among current female smokers only 18–25 % quit smoking after recognition of their pregnancy (‘‘The 2004 [5] united states surgeon general’s report: the health consequences of smoking,’’ 2004). Since the inclusion of maternal smoking history on US birth certificates beginning in 2003, data from 24 states indicate that 9.7 % of birth mothers overall (15.5 % non-Hispanic white, 8.7 % non-Hispanic Black, and 2.1 % Hispanic) smoked during pregnancy [6]. Furthermore, during 2000–2008, the prevalence of LBW infants increased from 7.6 to 8.2 %, while the prevalence of preterm also increased from 11.6 to 12.3 % [7].
The aim of our study was to evaluate the impact of smoking cessation during pregnancy and non-smoking on the prevalence of adverse infant outcomes (LBW and preterm) in San Bernardino County (SBC), California. SBC is the largest county in the contiguous U.S. and is home to over 2 million residents. Rankings based on an analysis by the Robert Wood Johnson Foundation of 3,000 counties within the US place SBC at 45th (of 56) in California based on health outcomes, and 50th on health factors [8]. As of 2009, SBC experienced one of the highest rates of infant mortality [7.5 deaths per 1,000 live births (‘‘Vital statistics query system,’’)], low-birth weight (7.1 % [10]) and pre-term births (11.5 % [10]) in the state. When considering county-level resources to address the burden of adverse infant outcomes, it is noteworthy that SBC faces severe economic challenges from foreclosures [11], municipal bankruptcy [12], and a high prevalence of unemployment and working poor [8, 13]. Our specific aims were:
To describe the association between never-smoking and LBW or pre-term outcomes among SBC birth mothers,
To describe the association between smoking cessation at pregnancy recognition and LBW and pre-term births among SBC birth mothers,
To examine the public health impact (i.e., number of adverse birth outcomes avoided) of smoking cessation at pregnancy recognition among SBC birth mothers.
These aims provide immediate, usable data for allocating resources to prevent maternal smoking in the county and the region.
Methods
Data Source
As a part of demographic analyses for the National Children’s Study, de-indentified Birth Cohort Files were obtained from California’s Department of Public Health (CDPH) for the years spanning 2002–2008 by Loma Linda University and University of California, Irvine. These birth files document 100 % of live births registered in California during each calendar year and contain information related to maternal and paternal demographic characteristics. After the implementation of the National Center for Health Statistics (NCHS) 2003 revision of the U.S. Standard Certificate of Live Birth in 2007, information on tobacco use during pregnancy became available to California (California Health and Safety Code section 102426). Therefore, maternal demographic and behavioral variables for all birth certificates recorded in SBC for 2007 and 2008 were extracted for analysis. Ethical review for this study was obtained by the Institutional Review Boards of University of California, Irvine and Loma Linda University.
Maternal Smoking Assessment
California implemented the 2003 revision of the birth certificate in 2007, and has since collected data from mothers on the number of cigarettes or packs of cigarettes smoked before and during pregnancy. Four items are recorded representing the number of cigarettes used in the 3 months before, as well as the first, second, and third trimester of pregnancy. The information is recorded by the facility of birth based on information obtained from the mother; or, if the birth did not occur in a facility, is completed by the attendant or certifier based on information obtained from the mother. For our analysis, women who reported smoking any number of cigarettes during any trimester of pregnancy were considered ‘‘smokers during pregnancy’’ while those who smoked in the 3 months before pregnancy but not in any of the three trimesters were considered ‘‘smoking cessation during pregnancy.’’
Adverse Infant Outcomes
Of the adverse infant outcomes reported on birth certificates, our analysis examined low birth weight and pre-term birth. Although the birth certificate includes variables on complications of pregnancy and delivery that report low birth weight or pre-term birth outcomes, we wanted to use a standardized method of outcome assessment and thus specifically coded these outcomes based on the actual birth weight and length of gestation recorded at delivery. Those births with a birth weight under 2,500 grams were classified as ‘‘low birth weight’’ and those with a length of gestation under 37 weeks (259 days) were classified as ‘‘pre-term birth’’ based on the U.S. national reference for fetal growth [14].
Statistical Analysis
Logistic Regression Analysis
To examine the relation between maternal smoking and adverse birth outcomes we conducted a multivariable logistic regression analysis. The outcome variables for the analysis were low birth weight and pre-term birth. The exposure variables were defined as indicators for smoking cessation during pregnancy, non-smoking (before or during pregnancy), and, as a referent, smoking during pregnancy. We constructed larger multivariable models by adding the following variables to the model and examining the change in exposure estimate: maternal age (0- < 18, 18–34, and 35 years or older), mother’s race/ethnicity (Hispanic, non-Hispanic Asian/Pacific Islander, non-Hispanic black, Hispanic, non-Hispanic white, and other/multi-ethnic/not-specified), mother’s years of education (0–8, 9–11, 12, 13–15, and 16 years or more), mother’s use of WIC services (yes vs. no), trimester prenatal care began (none, first, second, or third trimester), and the principal payer for prenatal care. Since many of these variables measured dimensions of the same potential confounder (i.e., socioeconomic status), we constructed models with confounders of age and one additional covariate.
Exposure Impact Number
To examine how smoking cessation during pregnancy or non-smoking (before and during pregnancy) can alleviate the burden of adverse infant outcomes in SBC, we conducted analyses to measure the number of the adverse infant outcomes that could be prevented by non-smoking or by smoking cessation. We note here that, in clinical trials, the number needed to treat (NNT) is often used in univariate analyses to calculate the number of adverse events that could be avoided by treatment [15]. In prospective observational studies, however, NNT can be limited by the need to control for confounding when examining an exposure effect [16, 17].
Therefore, to control for confounding when examining the effect of avoiding maternal smoking (i.e., by nonsmoking or smoking cessation) we calculated an exposure impact number (EIN) proposed by Bender et al. [16]. The EIN computes the number of events that could be avoided by removing an exposure and the estimate is based on an averaged risk difference (derived from a logistic regression model with exposure and confounders), where the risk difference for removal of exposure is averaged over the observed confounder values. Specifically, from the logistic model and expected risk difference (ERD) adjusted for confounders is given by,
where (x1,…, xk) represent binary or continuous confounders and π (x1…xk, z) represents the risk for the exposed (z = 1) and unexposed (z = 0) groups. An EIN is then given by 1/ERD.
A confidence interval for the EIN is produced by the inverse of the variance estimator for the ARD given by B · C · B‘ where B is the vector of the coefficients from the logistic model and C is the covariance matrix. All analyses were performed using SAS version 9.3 (SAS Institute Inc, Cary, North Carolina).
Results
There were 33,193 total live births in SBC in 2007, and 32,035 in 2008. Of those born in 2007, there were 1,430 children born from mothers who smoked during pregnancy, and 1,843 LBW and 3,480 pre-term deliveries. In 2008, there were 1,355 children born from mothers who smoked during pregnancy, 1,798 LBW and 3,238 preterm births. Tables 1 and 2 provide the demographic characteristics of infants born with LBW or preterm outcomes, respectively, for 2007 and 2008.
Table 1.
Frequency and percentages of selected characteristics of mothers of infants born with low birth weight or normal birth weight outcomes in San Bernardino County in 2007, 2008
| 2007 | 2008 | |||||
|---|---|---|---|---|---|---|
| Total 33,193 |
Normal weight 28,703 (94.0) |
Low birthweight 1,841 (6.0) |
Total 32,035 |
Normal weight 27,879 (94.0) |
Low birthweight 1,795 (6.0) |
|
| Maternal tobacco use | ||||||
| Never smoker | 28,547 (93.7) | 26,877 (93.9) | 1,670 (90.9) | 27,745 (93.6) | 26,130 (93.9) | 1,615 (90.2) |
| Smoker during pregnancy | 1,354 (4.4) | 1,220 (4.3) | 134 (7.3) | 1,284 (4.3) | 1,153 (4.1) | 131 (7.3) |
| Smoking cessation during pregnancy | 563 (1.9) | 530 (1.9) | 33 (1.8) | 605 (2.0) | 560 (2.0) | 45 (2.5) |
| Maternal age | ||||||
| <18 years | 1,365 (4.5) | 1,269 (4.4) | 96 (5.2) | 1,273 (4.3) | 1,169 (4.2) | 104 (5.8) |
| 18–<35 years | 25,712 (84.2) | 24,214 (84.4) | 1,498 (81.3) | 24,932 (84.0) | 23,517 (84.4) | 1,415 (78.7) |
| 35 years or older | 3,468 (11.4) | 3,219 (11.2) | 249 (13.5) | 3,472 (11.7) | 3,193 (11.5) | 279 (15.5) |
| Mother’s race/ethnicity | ||||||
| Hispanic | 18,421 (60.3) | 17,394 (60.6) | 1,027 (55.7) | 17,839 (60.1) | 16,817 (60.3) | 1,022 (56.8) |
| Non-Hispanic white | 7,461 (24.4) | 7,043 (24.5) | 418 (22.7) | 7,259 (24.5) | 6,867 (24.6) | 392 (21.8) |
| Non-Hispanic black | 2,669 (8.7) | 2,409 (8.4) | 260 (14.1) | 2,622 (8.8) | 2,374 (8.5) | 248 (13.8) |
| Asian/Pacific islander | 1,822 (6.0) | 1,693 (5.9) | 129 (7.0) | 1,748 (5.9) | 1,627 (5.8) | 121 (6.7) |
| Other/multi/unknown | 173 (0.6) | 164 (0.6) | 9 (0.5) | 209 (0.7) | 194 (0.7) | 15 (0.8) |
| Mother’s years of education | ||||||
| 0–8 years | 2,157 (7.1) | 2,029 (7.1) | 128 (7.0) | 1,880 (6.4) | 1,776 (6.4) | 104 (5.9) |
| 9–11 years | 7,284 (24.0) | 6,826 (24.0) | 458 (25.0) | 7,025 (23.9) | 6,548 (23.7) | 477 (26.9) |
| 12 years | 10,320 (34.1) | 9,677 (34.0) | 643 (35.0) | 9,627 (32.7) | 9,057 (32.8) | 570 (32.1) |
| 13–15 years | 9,425 (31.1) | 8,880 (31.2) | 545 (29.7) | 9,693 (33.0) | 9,142 (33.1) | 551 (31.1) |
| 16 years or more | 1,125 (3.7) | 1,063 (3.7) | 62 (3.4) | 1,186 (4.0) | 1,114 (4.0) | 72 (4.1) |
| Mother’s use of WIC | ||||||
| No | 12,841 (42.4) | 12,049 (42.4) | 792 (43.7) | 11,333 (38.4) | 10,581 (38.2) | 752 (42.2) |
| Yes | 17,426 (57.6) | 16,404 (57.7) | 1,022 (56.3) | 18,183 (61.6) | 17,153 (61.9) | 1,030 (57.8) |
| Trimester prenatal care began | ||||||
| No prenatal care | 271 (0.9) | 229 (0.8) | 42 (2.3) | 220 (0.8) | 184 (0.7) | 36 (2.0) |
| First trimester | 24,433 (80.6) | 22,986 (80.7) | 1,447 (79.3) | 23,963 (81.4) | 22,567 (81.6) | 1,396 (78.3) |
| Second trimester | 4,729 (15.6) | 4,441 (15.6) | 288 (15.8) | 4,476 (15.2) | 4,161 (15.1) | 315 (17.7) |
| Third trimester | 880 (2.9) | 832 (2.9) | 48 (2.6) | 781 (2.7) | 745 (2.7) | 36 (2.0) |
| Principal source of payment for prenatal care | ||||||
| Uninsured | 1,009 (3.3) | 904 (3.2) | 105 (5.7) | 825 (2.8) | 738 (2.7) | 87 (4.8) |
| Private insurance | 12,670 (41.5) | 11,955 (41.7) | 715 (38.8) | 12,171 (41.0) | 11,466 (41.1) | 705 (39.2) |
| Medi-Cal | 15,053 (49.3) | 14,122 (49.2) | 931 (50.5) | 14,793 (49.9) | 13,877 (49.8) | 916 (51.0) |
| Other | 1,814 (5.9) | 1,722 (6.0) | 92 (5.0) | 1,888 (6.4) | 1,798 (6.5) | 90 (5.0) |
Low birth weight defined as a recorded birth weight of <2,500 g on the registered birth certificate
Table 2.
Frequency and percentages of selected characteristics of mothers of infants born with pre-term or normal term outcomes in San Bernardino County in 2007, 2008
| 2007 | 2008 | |||||
|---|---|---|---|---|---|---|
| Total 33,193 |
Normal term 29,713 (89.5) |
Preterm 3,480 (10.5) |
Total 32,035 |
Normal term 28,797 (89.9) |
Preterm 3,238 (10.1) |
|
| Maternal tobacco use | ||||||
| Never smoker | 31,050 (93.8) | 27,857 (94.0) | 3,193 (92.2) | 29,989 (93.7) | 27,011 (93.9) | 2,978 (92.2) |
| Smoker during pregnancy | 1,430 (4.3) | 1,224 (4.1) | 206 (6.0) | 1,355 (4.2) | 1,168 (4.1) | 187 (5.8) |
| Smoking cessation during pregnancy | 626 (1.9) | 562 (1.9) | 64 (1.9) | 648 (2.0) | 584 (2.0) | 64 (2.0) |
| Maternal age | ||||||
| <18 years | 1,422 (4.3) | 1,242 (4.2) | 180 (5.2) | 1,317 (4.1) | 1,158 (4.0) | 159 (4.9) |
| 18–<35 years | 27,918 (84.1) | 25,115 (84.5) | 2,803 (80.6) | 26,886 (83.9) | 24,286 (84.3) | 2,600 (80.3) |
| 35 years or older | 3,851 (11.6) | 3,355 (11.3) | 496 (14.3) | 3,832 (12.0) | 3,353 (11.6) | 479 (14.8) |
| Mother’s race/ethnicity | ||||||
| Hispanic | 20,059 (60.4) | 17,913 (60.3) | 2,146 (61.7) | 19,213 (60.0) | 17,251 (59.9) | 1,962 (60.6) |
| Non-Hispanic white | 8,218 (24.8) | 7,489 (25.2) | 729 (21.0) | 8,013 (25.0) | 7,275 (25.3) | 738 (22.8) |
| Non-Hispanic black | 2,800 (8.4) | 2,417 (8.1) | 383 (11.0) | 2,756 (8.6) | 2,417 (8.4) | 339 (10.5) |
| Asian/Pacific islander | 1,925 (5.8) | 1,721 (5.8) | 204 (5.9) | 1,824 (5.7) | 1,650 (5.7) | 174 (5.4) |
| Other/multi/unknown | 191 (0.6) | 173 (0.6) | 18 (0.5) | 229 (0.7) | 204 (0.7) | 25 (0.8) |
| Mother’s years of education | ||||||
| 0–8 years | 2,360 (7.2) | 2,075 (7.0) | 285 (8.3) | 2,045 (6.4) | 1,805 (6.3) | 240 (7.5) |
| 9–11 years | 7,831 (23.8) | 6,889 (23.4) | 942 (27.3) | 7,504 (23.6) | 6,633 (23.2) | 871 (27.2) |
| 12 years | 11,145 (33.8) | 9,974 (33.8) | 1,171 (33.9) | 10,399 (32.8) | 9,340 (32.7) | 1,059 (33.1) |
| 13–15 years | 10,390 (31.6) | 9,444 (32.0) | 946 (27.4) | 10,522 (33.1) | 9,612 (33.7) | 910 (28.4) |
| 16 years or more | 1,208 (3.7) | 1,099 (3.7) | 109 (3.2) | 1,279 (4.0) | 1,157 (4.1) | 122 (3.8) |
| Mother’s use of WIC | ||||||
| No | 14,045 (42.7) | 12,634 (42.9) | 1,411 (41.0) | 12,328 (38.7) | 11,089 (38.7) | 1,239 (38.5) |
| Yes | 18,851 (57.3) | 16,824 (57.1) | 2,027 (59.0) | 19,535 (61.3) | 17,557 (61.3) | 1,978 (61.5) |
| Trimester prenatal care began | ||||||
| No prenatal care | 289 (0.9) | 199 (0.7) | 90 (2.6) | 230 (0.7) | 162 (0.6) | 68 (2.1) |
| First trimester | 26,630 (80.8) | 23,992 (81.3) | 2,638 (76.7) | 25,949 (81.6) | 23,468 (82.1) | 2,481 (77.5) |
| Second trimester | 5,084 (15.4) | 4,480 (15.2) | 604 (17.6) | 4,780 (15.0) | 4,221 (14.8) | 559 (17.5) |
| Third trimester | 941 (2.9) | 832 (2.8) | 109 (3.2) | 826 (2.6) | 734 (2.6) | 92 (2.9) |
| Principal source of payment for prenatal care | ||||||
| Uninsured | 1,084 (3.3) | 878 (3.0) | 206 (5.9) | 888 (2.8) | 726 (2.5) | 162 (5.0) |
| Private insurance | 13,920 (41.9) | 12,647 (42.6) | 1,273 (36.6) | 13,295 (41.5) | 12,099 (42.0) | 1,196 (36.9) |
| Medi-cal | 16,236 (48.9) | 14,420 (48.5) | 1,816 (52.2) | 15,844 (49.5) | 14,142 (49.1) | 1,702 (52.6) |
| Other | 1,953 (5.9) | 1,768 (6.0) | 185 (5.3) | 2,008 (6.3) | 1,830 (6.4) | 178 (5.5) |
Pre-term birth is defined as a recorded length of gestation of <37 weeks (259 days) on the registered birth certificate
Table 3 presents the age-adjusted univariate odds ratios for LBW and preterm outcomes for each of the selected characteristics of mothers in SBC in 2007 and 2008. Compared to mothers who smoked during pregnancy, nonsmoking mothers had a substantially lower risk for LBW [OR: 0.56 (95 % CI 0.47, 0.68), 2007; 0.54 (0.44, 0.65), 2008] and preterm outcomes [0.68 (0.58, 0.79), 2007; 0.68 (0.58, 0.80), 2008] in both calendar years. Similarly, mothers who quit smoking during the pregnancy had reduced risks of LBW [0.57 (0.39, 0.85), 2007; 0.72 (0.50, 1.02), 2008] and preterm outcomes [0.69 (0.51, 0.92), 2007; 0.69 (0.51, 0.93), 2008] compared to those who continued to smoke during pregnancy. Non-Hispanic black and Asian/Pacific Islander race/ethnicity of mother was shown to have increased risk for LBW and preterm outcomes when compared to non-Hispanic white mothers. Hispanic race/ethnicity was also shown to have increased risk for preterm outcomes in each calendar year when compared to non-Hispanic white mothers [1.22 (1.12, 1.34), 2007; 1.12 (1.02, 1.22), 2008], but the association for LBW was not as clear. More than 12 years of education demonstrated significant reductions in risk for preterm birth at any level, while less than 12 years of education showed significant increase in risk for preterm birth at any level. Use of WIC services was shown to have a slight reduction in risk for LBW [0.95 (0.87, 1.05), 2007; 0.85 (0.77, 0.94), 2008] despite slightly increasing the risk for preterm outcomes [1.09 (1.01, 1.17), 2007; 1.02 (0.95, 1.10), 2008]. Interestingly, mothers who began their prenatal care at any trimester had a significant reduction in risk for LBW and preterm births in either year observed compared to mothers who had no prenatal care whatsoever. Lack of insurance, and Medi-Cal as a primary payer for Prenatal care showed increased risk for LBW and preterm outcomes when compared to those with a private insurance payer.
Table 3.
Age-adjusted univariate odds ratios for adverse birth outcomes for selected characteristics of mothers in San Bernardino County in 2007, 2008
| Low birth weight | Pre-term birth | |||
|---|---|---|---|---|
| 2007 N = 30,546 |
2008 N = 29,677 |
2007 N = 33,193 |
2008 N = 32,035 |
|
| Maternal tobacco use | ||||
| Never smoker | 0.56 (0.47, 0.68) | 0.54 (0.44, 0.65) | 0.68 (0.58, 0.79) | 0.68 (0.58, 0.80) |
| Smoker during pregnancy | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Smoking cessation during pregnancy | 0.57 (0.39, 0.85) | 0.72 (0.50, 1.02) | 0.69 (0.51, 0.92) | 0.69 (0.51, 0.93) |
| Maternal age | ||||
| <18 years | 1.22 (0.99, 1.52) | 1.48 (1.20, 1.82) | 1.30 (1.11, 1.53) | 1.28 (1.08, 1.52) |
| 18–<35 years | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| 35 years or older | 1.25 (1.09, 1.44) | 1.45 (1.27, 1.66) | 1.33 (1.20, 1.47) | 1.34 (1.20, 1.48) |
| Mother’s race/ethnicity | ||||
| Hispanic | 0.99 (0.88, 1.11) | 1.05 (0.94, 1.19) | 1.22 (1.12, 1.34) | 1.12 (1.02, 1.22) |
| Non-Hispanic white | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Non-Hispanic black | 1.81 (1.54, 2.13) | 1.82 (1.54, 2.15) | 1.62 (1.42, 1.84) | 1.38 (1.20, 1.58) |
| Asian/Pacific islander | 1.26 (1.03, 1.54) | 1.24 (1.01, 1.54) | 1.18 (1.00, 1.40) | 1.00 (0.84, 1.19) |
| Other/multi/unknown | 0.91 (0.46, 1.80) | 1.33 (0.78, 2.27) | 0.99 (0.60, 1.63) | 1.19 (0.78, 1.81) |
| Mother’s years of education | ||||
| 0–8 years | 0.92 (0.76, 1.12) | 0.87 (0.70, 1.08) | 1.13 (0.98, 1.30) | 1.12 (0.96, 1.30) |
| 9–11 years | 0.98 (0.86, 1.12) | 1.10 (0.97, 1.26) | 1.15 (1.04, 1.26) | 1.14 (1.03, 1.26) |
| 12 years | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| 13–15 years | 0.91 (0.81, 1.03) | 0.93 (0.83, 1.05) | 0.84 (0.77, 0.92) | 0.82 (0.75, 0.90) |
| 16 years or more | 0.84 (0.64, 1.10) | 0.95 (0.73, 1.22) | 0.80 (0.65, 0.98) | 0.87 (0.72, 1.07) |
| Mother’s use of WIC | ||||
| No | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Yes | 0.95 (0.87, 1.05) | 0.85 (0.77, 0.94) | 1.09 (1.01, 1.17) | 1.02 (0.95, 1.10) |
| Trimester prenatal care began | ||||
| No prenatal care | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| First trimester | 0.34 (0.25, 0.48) | 0.32 (0.22, 0.46) | 0.24 (0.19, 0.31) | 0.25 (0.19, 0.34) |
| Second trimester | 0.35 (0.25, 0.50) | 0.39 (0.27, 0.57) | 0.30 (0.23, 0.39) | 0.32 (0.24, 0.43) |
| Third trimester | 0.31 (0.20, 0.49) | 0.25 (0.15, 0.40) | 0.29 (0.21, 0.39) | 0.30 (0.21, 0.43) |
| Primary source of payment for prenatal care | ||||
| Private insurance | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) | 1.00 (Reference) |
| Medi-Cal | 1.12 (1.01, 1.24) | 1.09 (0.99, 1.21) | 1.27 (1.18, 1.37) | 1.24 (1.15, 1.34) |
| Other | 0.91 (0.73, 1.14) | 0.84 (0.67, 1.05) | 1.06 (0.90, 1.25) | 1.01 (0.86, 1.19) |
| Uninsured | 1.95 (1.57, 2.42) | 1.92 (1.52, 2.43) | 2.35 (1.99, 2.76) | 2.26 (1.89, 2.71) |
Table 4 presents multivariable odds ratios relating maternal smoking to LBW and preterm outcomes, respectively, with adjustment for age and one additional covariate for 2007 and 2008. Non-smoking mothers showed substantially lower risk for LBW and preterm outcomes in each calendar year compared to those who smoked during pregnancy. In 2007, smoking cessation during pregnancy was associated with reduced risk for LBW when adjusting or age, and any of the added covariates. This association was not significant in 2008, however, despite point estimates demonstrating reduced odds ratios compared to those mothers who smoked during pregnancy. In 2007 and 2008, the risk for preterm outcome was significantly reduced when adjusting for age and mother’s race/ethnicity or use of WIC services.
Table 4.
Multivariable odds ratios relating maternal smoking to low birth weight and pre-term birth with adjustment for age and one additional covariate among mothers in San Bernardino County, 2007–2008
| 2007 | 2008 | |||||
|---|---|---|---|---|---|---|
| Never smoker | Smoker during pregnancy |
Smoking cessation during pregnancy |
Never smoker | Smoker during pregnancy |
Smoking cessation during pregnancy |
|
| n = 31,050 | n = 1,430 | n = 626 | n = 29,989 | n = 1,355 | n = 648 | |
| Low birth weight | ||||||
| Age | 0.56 (0.47, 0.68) | 1.00 (Reference) | 0.57 (0.39, 0.85) | 0.54 (0.44, 0.65) | 1.00 (Reference) | 0.72 (0.50, 1.02) |
| Age + ethnicity | 0.57 (0.47, 0.69) | 1.00 (Reference) | 0.58 (0.39, 0.86) | 0.53 (0.43, 0.64) | 1.00 (Reference) | 0.72 (0.51, 1.03) |
| Age + education | 0.59 (0.49, 0.71) | 1.00 (Reference) | 0.59 (0.40, 0.88) | 0.55 (0.45, 0.66) | 1.00 (Reference) | 0.73 (0.51, 1.04) |
| Age + WIC | 0.55 (0.46, 0.67) | 1.00 (Reference) | 0.57 (0.38, 0.85) | 0.55 (0.45, 0.66) | 1.00 (Reference) | 0.73 (0.51, 1.03) |
| Age + trimester prenatal care began |
0.60 (0.49, 0.72) | 1.00 (Reference) | 0.61 (0.41, 0.91) | 0.56 (0.46, 0.68) | 1.00 (Reference) | 0.73 (0.51, 1.04) |
| Age + primary source of payment for prenatal care | 0.60 (0.49, 0.72) | 1.00 (Reference) | 0.62 (0.42, 0.93) | 0.55 (0.45, 0.67) | 1.00 (Reference) | 0.76 (0.53, 1.08) |
| Pre-term birth | ||||||
| Age | 0.68 (0.58, 0.79) | 1.00 (Reference) | 0.69 (0.51, 0.92) | 0.68 (0.58, 0.80) | 1.00 (Reference) | 0.69 (0.51, 0.93) |
| Age + ethnicity | 0.62 (0.53, 0.73) | 1.00 (Reference) | 0.68 (0.50, 0.91) | 0.66 (0.56, 0.77) | 1.00 (Reference) | 0.69 (0.51, 0.93) |
| Age + education | 0.72 (0.61, 0.84) | 1.00 (Reference) | 0.73 (0.54, 0.98) | 0.73 (0.54, 0.98) | 1.00 (Reference) | 0.73 (0.54, 0.98) |
| Age + WIC | 0.67 (0.58, 0.78) | 1.00 (Reference) | 0.69 (0.51, 0.93) | 0.69 (0.58, 0.80) | 1.00 (Reference) | 0.69 (0.51, 0.93) |
| Age + trimester prenatal care began | 0.76 (0.65, 0.89) | 1.00 (Reference) | 0.75 (0.55, 1.02) | 0.75 (0.63, 0.88) | 1.00 (Reference) | 0.74 (0.55, 1.01) |
| Age + primary source of payment for prenatal care | 0.75 (0.64, 0.88) | 1.00 (Reference) | 0.77 (0.57, 1.04) | 0.74 (0.63, 0.87) | 1.00 (Reference) | 0.76 (0.56, 1.02) |
To examine the public health impact of smoking cessation at pregnancy recognition, we calculated the EIN for LBW [EIN: 35.3 (95 % CI 21.1, 108.2)] and preterm utcomes [EIN: 28.3 (95 % CI 17.6, 72.4)]. These results indicate that, on average, 35–36 mothers who smoked during pregnancy could have avoided one case of LBW by ceasing to smoke during their pregnancy. Due to estimation uncertainty the EIN may also lie between 21 and 109 mothers ceasing to smoke during pregnancy to prevent LBW outcome in one additional child compared to mother’s who continue to smoke during pregnancy. Similarly, 28–29 mothers who smoked during pregnancy could have avoided one case of preterm birth by ceasing to smoke during their pregnancy with an uncertainty ranging from 17 to 73 mothers. We also calculated the EIN for non-smoking relative to smoking during pregnancy; the public health impact of non-smoking was markedly greater, with EINs for LBW [EIN: 24.5 (95 % CI 19.0, 34.3)] and preterm outcomes [EIN: 24.5 (95 % CI 18.5, 36.1)].
Discussion
Our major findings are as follows: (1) relative to maternal smokers, a significantly lower risk of low birth weight was found for non-smoking mothers and for mothers who quit smoking during pregnancy; (2) relative to maternal smokers, a significantly lower risk of pre-term birth was found for non-smoking mothers and for mothers who quit smoking during pregnancy; (3) an exposure impact assessment indicating a single low birth-weight or pre-term birth in the county could be prevented either by 35 mothers quitting smoking during pregnancy or by 25 mothers being pre-pregnancy non-smokers.
Our findings identify a strong etiologic link between maternal smoking and adverse infant outcomes (LBW, pre-term births) in San Bernardino County-an ethnically diverse county with high rates of adverse infant outcomes and health care resources severely limited by economic downturn. Our approach involved the use of novel methodology to provide an estimate of how many women are needed to quit smoking during pregnancy or avoid smoking pre-pregnancy (i.e., by cessation or lifetime never-smoking) in order to prevent adverse infant outcomes that are often costly in terms of extended stays in neonatal intensive care units (NICUs) [18, 19] and the long term, untold, deficits experienced by surviving infants [20, 21].
Implications of Allocating Resources to Smoking Cessation Programs Designed for Pregnant Women
Our analysis indicated that one low birth weight or pre-term birth outcome could be prevented either by 35 women quitting smoking during pregnancy or by 22 women quitting/avoiding smoking pre-pregnancy. These data immediately translate into modeling the effectiveness of existing programs and allocating limited resources to new programs.
One implication of our findings is that it is cost-effective to incorporate cessation services specific to all pregnant women in San Bernardino County. A cost-benefit/cost-effectiveness analysis conducted by Marks et al. [18] found that a program offered to all pregnant smokers would shift 5,876 LBW infants to normal weight and would cost about $4,000 for each LBW infant prevented. Using the same parameters of their study ($30 per participant and 15 % of the participants would quit smoking), paired with the EIN values we estimated, we estimate that it would cost about $7,000 for each LBW or preterm infant prevented.
When considering the costs associated with NICU treatment, the costs of adverse infant outcomes far exceed the costs of smoking cessation efforts. For example, daily NICU costs exceeding $3,500 per infant are not uncommon [19] and in an analysis of 680 newborns from the Healthcare Cost and Utilization Project Kids Inpatient Discharge (HCUP-KIDs) data, LBW infants typically stayed for 228 days with average costs of $703,356 (SD $19,846) [19]. In the cost-benefit analysis mentioned previously, Marks et al. [18] estimated that smoking cessation programs would save $77,807,054, or $3.31 per $1 spent compared to the cost of treating LBW infants in a NICU. Even further, the ratio of savings to cost increase to more than six to one when they included long-term care for infants with disabilities secondary to LBW [18].
A second implication of our findings is that San Bernardino County is lacking in smoking cessation efforts targeted towards pregnant women. The 2001 Surgeon General’s Report on Women and Smoking stated that 20–25 % of women of lower socioeconomic status enrolled in intensive smoking cessation programs were successful in smoking cessation, but only a small proportion take advantage of such programs (2001). It could be the case that women are not taking advantage of these programs because of a lack of awareness, in which case allocating resources to promote these programs would help improve their utilization. Alternatively, there could also be a lack of program availability. This latter possibility seems to be the case in San Bernardino County. According to the California Smoker’s Helpline, San Bernardino County currently has only one program, the Comprehensive Tobacco Treatment Program (CTTP), that offers smoking cessation services designed specifically for women. The expansion of smoking cessation programs specific to pregnant women or the inclusion of services specific to pregnant women among other existing cessation programs that do not address pregnant women could provide an avenue for substantial improvement not only in the number of women who quit during pregnancy but also in reducing the incidence of adverse infant outcomes.
Strengths and Limitations
To our knowledge, these are the first county-wide maternal tobacco use findings from San Bernardino County, one of the most diverse counties in the United States, and the first use of state birth certificates for this purpose. Furthermore, our findings not only confirm but also attempt to quantify the direct impact of prenatal smoking cessation on the prevention of adverse infant outcomes. Results of this study are not generalizable to the entire state of California or the country as a whole because they are not a random sample of all births. Additionally, these results may underestimate the prevalence of tobacco use during pregnancy because underreporting of cigarette smoking is higher among pregnant smokers than among non-pregnant smokers [22].
Conclusion
We estimated that 35 women are needed to stop smoking during pregnancy to prevent one excess case of LBW or preterm birth in SBC. Our findings are concordant with the convincing data linking maternal smoking to LBW and preterm outcomes in the U.S. and our impact calculations provide immediate implications for allocating resources to maternal smoking cessation and prevention programs in a low resource county.
Acknowledgments
This study was supported by Grant HHSN 267200700021C from NICHD/Department of Health and Human Services (National Children’s Study Award to University of California, Irvine, and sub-award to Loma Linda University and Cal State San Bernardino).
Contributor Information
Michael Batech, Email: mbatech@llu.edu, Loma Linda University School of Public Health, Loma Linda, CA, USA; Loma Linda University Center for Health Research, 24951 North Circle Drive, Nichol Hall 1517, Loma Linda, CA 92350, USA.
Serena Tonstad, Email: stonstad@llu.edu, Department of Health Promotion and Education, Loma Linda University School of Public Health, Loma Linda, CA, USA.
Jayakaran S. Job, Email: jjob@llu.edu, Department of Global Health, Loma Linda University School of Public Health, Loma Linda, CA, USA.
Richard Chinnock, Email: rchinnock@llu.edu, Department of Pediatrics, Loma Linda University Medical Center, Loma Linda, CA, USA.
Bryan Oshiro, Email: boshiro@llu.edu, Department of Gynecology and Obstetrics, Loma Linda University Medical Center, Loma Linda, CA, USA.
T. Allen Merritt, Email: tamerritt@llu.edu, Department of Pediatrics, Loma Linda University School of Medicine, Loma Linda, CA, USA.
Gretchen Page, Email: gpage@llu.edu, Southern Inland Counties Regional Perinatal Program, Loma Linda University Medical Center, Loma Linda, CA, USA.
Pramil N. Singh, Email: psingh@llu.edu, Loma Linda University School of Public Health, Loma Linda, CA, USA; Loma Linda University Center for Health Research, 24951 North Circle Drive, Nichol Hall 1517, Loma Linda, CA 92350, USA.
References
- 1.Culyer AJ, Newhouse JP. Handbook of health economics. 1st ed. Amsterdam, New York: Elsevier; 2000. [Google Scholar]
- 2.Cnattingius S. The epidemiology of smoking during pregnancy: Smoking prevalence, maternal characteristics, and pregnancy outcomes. Nicotine & Tobacco Research. 2004;6(Suppl. 2):S125–S140. doi: 10.1080/14622200410001669187. [DOI] [PubMed] [Google Scholar]
- 3.Women and smoking: A report of the surgeon general. Atlanta (GA): 2001. [Google Scholar]
- 4.Centers for Disease, Control, and Prevention. Smoking prevalence among women of reproductive age—United States, 2006. Morbidity and Mortality Weekly Report. 2008;57(31):849–852. [PubMed] [Google Scholar]
- 5.The 2004 United States surgeon general’s report: The health consequences of smoking. N S W Public Health Bull. 2004;15(5–6):107. [PubMed] [Google Scholar]
- 6.Osterman MJ, Martin JA, Mathews TJ, Hamilton BE. Expanded data from the new birth certificate, 2008. National Vital Statistics Reports. 2011;59(7):1–28. [PubMed] [Google Scholar]
- 7.Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Mathews TJ, Osterman MJ. Births: Final data for 2008. National Vital Statistics Reports. 2010;59(1):3–71. [PubMed] [Google Scholar]
- 8.Institute, University of Wisconsin Population Health. County health rankings 2012. 2012. [Google Scholar]
- 9.Vital Statistics Query System. State of California, Department of Public Health, Birth and Death Records. 2010 http://www.apps.cdph.ca.gov/vsq/screen_idra.asp?cnty_cd=36&Year_Data=2009&Stats=4.
- 10.CDPH. Vital statistics query system. Retrieved 16 July 2012 from http://www.applications.dhs.ca.gov/vsq/default.asp.
- 11.Knox N. Foreclosures plague san bernardino county, calif., USA TODAY. 2008 Retrieved from http://www.usatoday.com/money/economy/housing/closetohome/2008-01-28-san-bernardino_N.htm.
- 12.Willon P. Bankrupt san bernardino must cut spending by a third, Los Angeles Times. 2012 Jul 25; 2012, Retrieved from http://www.latimes.com/news/local/la-me-san-bernardino-cuts-20120726,0,397850.story.
- 13.Shimura T. County unemployment rate increases despite job growth, High Desert Daily Press. Retrieved from 2012 http://www.vvdailypress.com/news/jobs-35672-rate-added.html.
- 14.Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A united states national reference for fetal growth. Obstetrics and Gynecology. 1996;87(2):163–168. doi: 10.1016/0029-7844(95)00386-X. [DOI] [PubMed] [Google Scholar]
- 15.Cook RJ, Sackett DL. The number needed to treat: A clinically useful measure of treatment effect. British Medical Journal. 1995;310(6977):452–454. doi: 10.1136/bmj.310.6977.452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bender R, Kuss O, Hildebrandt M, Gehrmann U. Estimating adjusted NNT measures in logistic regression analysis. Statistics in Medicine. 2007;26(30):5586–5595. doi: 10.1002/sim.3061. [DOI] [PubMed] [Google Scholar]
- 17.Heller RF, Dobson AJ, Attia J, Page J. Impact numbers: measures of risk factor impact on the whole population from case-control and cohort studies. Journal of Epidemiol Community Health. 2002;56(8):606–610. doi: 10.1136/jech.56.8.606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Marks JS, Koplan JP, Hogue CJ, Dalmat ME. A cost-benefit/cost-effectiveness analysis of smoking cessation for pregnant women. American Journal of Preventive Medicine. 1990;6(5):282–289. [PubMed] [Google Scholar]
- 19.Catlin A. Extremely long hospitalizations of newborns in the United States: data, descriptions, dilemmas. Journal of Perinatol. 2006;26(12):742–748. doi: 10.1038/sj.jp.7211617. [DOI] [PubMed] [Google Scholar]
- 20.Swamy GK, Ostbye T, Skjaerven R. Association of preterm birth with long-term survival, reproduction, and next-generation preterm birth. Journal of the American Medical Association. 2008;299(12):1429–1436. doi: 10.1001/jama.299.12.1429. [DOI] [PubMed] [Google Scholar]
- 21.Hack M, Taylor HG, Drotar D, et al. Chronic conditions, functional limitations, and special health care needs of school-aged children born with extremely low-birth-weight in the 1990s. Journal of the American Medical Association. 2005;294(3):318–325. doi: 10.1001/jama.294.3.318. [DOI] [PubMed] [Google Scholar]
- 22.Dietz PM, Homa D, England LJ, et al. Estimates of nondisclosure of cigarette smoking among pregnant and nonpreg-nant women of reproductive age in the United States. American Journal of Epidemiology. 2011;173(3):355–359. doi: 10.1093/aje/kwq381. [DOI] [PubMed] [Google Scholar]