ate the list of women to be sampled each month from the state birth certificate registry and the various methods used to promote completion of the surveys . Two sources of information comprise the PRAMS dataset – data from birth certificates and PRAMS surveys . In addition , the dataset includes weighting variables generated by CDC so that results can be generalized beyond the sampled women and to compare results across participating states . New York City and 39 states participate in the PRAMS protocol , including the Commonwealth of Pennsylvania that joined the PRAMS network in 2006 and began administering the surveys in 2007 . Based on a written request to the Pennsylvania Department of Health , Bureau of Health Statistics and Research , we were given access to the most current PRAMS data at the time of the project ( 2011 ) ( Pennsylvania Department of Health , 2014 ). The dataset was de-identified and thus included no distinguishing personal data or identifiers from the participating women . The University of Pittsburgh Institutional Review Board reviewed and approved the study .
The PA dataset included information from 1,080 postpartum women ( cases ). Because of large amounts of missing data ( up to 15 % per case ), 19 cases were removed leaving 1061 cases . Since the purpose of this study was to identify potential predictors of postpartum relapse to smoking , only women who reported smoking before their pregnancy and abstaining during the pregnancy were eligible resulting in 145 cases for this analysis .
Of the 319 variables included in the PRAMS dataset , 24 variables were selected for inclusion in the analysis based on relevant maternal characteristics and other potential risk factors for postpartum relapse reported in the current research literature . SPSS for Windows ( version 21 ) was used for the analysis . Missing data were replaced via expectation-maximization ( EM ) imputation . Logistic regression was used to identify meaningful predictors of postpartum relapse to smoking .
Results
Characteristics of the 145 women who were included in this study are shown in Table 1 . The sample was mainly white ( 79.8 %), first-time mothers ( 51.6 %), and infants born at full-term ( 92.2 %). Over half of the sample had 13 or more years of education . Approximately 50 % reported an unplanned pregnancy and one in four never attempted to breastfeed their newborn .
Table 1 . Sample characteristics (%, N = 145 ) ( Pennsylvania Department of Health , 2014 )
Race |
White |
79.8 |
Black |
13.2 |
Other |
6.9 |
Age |
24 years or younger |
41.4 |
25-29 years |
31.0 |
30-34 years |
17.2 |
35 years or older |
10.4 |
Education |
11 years or less |
15.2 |
12 years |
26.1 |
13-15 years |
37.0 |
16 years or more |
21.8 |
Marital status |
Married |
43.4 |
Other |
56.6 |
Total income during the year before pregnancy |
< $ 20,000 |
35.5 |
$ 20,000 - $ 50,000 |
33.4 |
> $ 50,000 |
31.1 |
WIC services used during pregnancy |
No |
50.9 |
Yes |
49.1 |
Previous live birth ( s ) |
0 |
51.6 |
1 |
29.5 |
2 |
13.7 |
≥ 3 |
5.2 |
Intentions related to pregnancy |
Planned |
34.8 |
Planned and desired sooner |
13.9 |
Unplanned |
51.2 |
Gestation length |
≤ 33 weeks |
2.7 |
34-36 weeks |
5.1 |
≥ 37 weeks |
92.2 |
Breastfeeding duration |
None |
26.1 |
0-6 weeks |
34.5 |
7-26 weeks |
14.2 |
Now |
25.2 |
Issue 2 2017 Pennsylvania Nurse 6