Antenatal care (ANC) has increased in coverage and improved in
efficiency. Yet, recent studies report differences across socioeconomic
groups (SES) in treatment of antenatal care and screening (Petersson
etal 2016; Rowe, Garcia & Davidson 2004; Smith et al 2016; Vassy,
Rosman & Rousseau 2014). Such disparities may through accessdifferences,
as regional care providers differ in policy or
implementation. Less educated women may be overrepresented in
hospital wards where a particular ANC treatment is less frequently
utilized (Petersson etal 2016). Alternatively, heterogeneous ANC
usage may origin from SES variability in predispositions towards
medical intervention during conception, or the ability to correctly
assess risks associated with treatment/non treatment. Previous
studies indicate that less education women (and their partners) are
more likely to reject ANC treatments when offered (Smith et al 2016).
No studies exist that draw from large scale representative samples that cover both offer and uptake of ANC (Rowe, Garcia & Davidson
2004). This project explore causes of ANC treatment-disparities
across SES by analyzing disparities in offer and/or uptake of ANC. We
focus on amniocenteses & placenty biopsy, relatively common ANC
treatments where previous research have reported socioeconomic
differences.
We ask two questions
Does the probability of being offered amniocentesis vary over
socioeconomic position?
Does the probability of going through with amniocentesis
treatment, if offered, vary over socioeconomic position?
We use the Norwegian Mother and Child Cohort Study (Magnus
et al 2016). A key strength of MoBa is that it allows us to
separately solicit mothers on being offered amniocentesis (N =
10387) and going through with the operation (N = 1514). Health
indicators of parent and fetus and socioeconomic indicators of
mothers and fathers will be used.
The main outcomes are (a) reported offer of amniocentesis and
(b) reporter treatment or not. The main exposure variables are
measures of SES (detailed information on variable construction
provided in protocol). We also plan to include in the analysis
anonymous hospital dummies, maternal general health status,
previous births, and current pregnancy treatments. We model
possible SES discrepancies as transitions – first considering
covariates for being offered treatment and second subjecting
oneself to treatment or not. We decompose the SES effect of
each transition.
The study will be the first to assess the findings of qualitative
and less extensive quantitative studies reporting SES
differences in ANC treatment, and the first to tentatively
differentiate the role of offer from the role of uptake. We believe
that this is important and beneficial for conducting informed
medical policy regarding amniocentesis in particular and ANC in
general. Due to the large sample size, the finding has potential
to be generalizable to other context and thus be of significance
to a wide audience of practitioners, researchers and policy
makers.
