We conducted a non-randomized controlled trial to investigate if women’s exposure
to TBA referrals and a community education program changed their medical facility/SBA-seeking
behavior for delivery. We hypothesized that enlisting the support of TBA referrals
would significantly increase the number of women who choose to give birth in a health
facility with a SBA present.
The site of the study is in southeastern Kenya, in Yatta sub-county of Machakos County.
The region is part of the arid and semi-arid lands of Kenya, populated by the Kamba
tribe. Dry land rain-fed agriculture and small-scale animal husbandry are the primary
source of livelihood for the non-urban population that is included in this study.
Extreme poverty and geographic isolation of the population limit access to health
care and contribute to a high child and neonatal mortality. The intervention facility
was chosen because of its remarkably low percentage (2.4Â %) of deliveries of antenatal
care patients occurring in a facility with a SBA 16]. The Yatta sub-county has a female population of 141,075 and a total population of
273,519 17]. A planning meeting was held with the Kisiiki Health Center staff in which the purpose
and methods of the intervention were presented. After gaining the support of the clinical
staff, they contacted TBAs known to them and recruited the assistance of the chief
and the local community health workers to contact other TBAs who were active in the
catchment area of the health center. The TBAs who were contacted were asked to attend
a meeting with the investigators to discuss the project and request their participation
and consent.
In February 2013, we held a meeting with TBAs from the area surrounding the Kisiiki
healthcare facility in Yatta sub-county to recruit them into the TBA referral program
in order to expand the work of Tomedi et al. 16]. We met with the TBAs and encouraged them to educate or inform their clients about
the potential complications that can occur for the mother and newborn at the time
of delivery (with examples solicited from experienced TBAs), and why it is important
to deliver at a health facility that can handle the complications or provide transport
to a higher level of care (e.g. blood transfusion, C-section). Enrolled TBAs were
told that they would be compensated with a per diem of KSH 200 (approximately $2.50
USD) for each pregnant woman that the TBAs referred to a facility for an SBA delivery.
The Kenya government/Ministry of Health has emphasized the importance of health facility
deliveries and that TBAs are not legitimate providers of health care (e.g. they are
practicing without a license). The TBAs are aware of that, but in this and other meetings
have stated that pregnant women continue to come to them requesting their assistance,
and they feel an obligation to help. The lead author was told in preliminary meetings
with TBAs (before the study was started) that they do not charge a fee for their services,
but other community sources (e.g. the chief of some of the villages, local health
staff) have said that they are paid by the client’s family, an amount that could range
from KSH 200 to 500. Because of the government’s stated position regarding TBAs, they
have been reluctant to refer the client when complications occur. The pregnant women,
when asked about barriers to facility birth, have expressed a fear that they might
deliver while traveling the long distance to a health facility. Therefore, in addition
to providing the TBAs with an educational message for them to deliver to their clients,
the meeting addressed other concerns and barriers. A Ministry of Health official in
charge of maternity services for the district spoke of the important role of the TBAs
to educate their clients and improve maternal outcomes. She and the hospital staff
welcomed and encouraged the involvement of TBAs (up to the point of delivery). It
was recognized that the TBAs would continue to be consulted by many families. They
are asked to assess if a woman is in active labor, and if there likely is adequate
time for the trip to the health facility. The TBA then accompanies the woman in case
delivery should occur en route. The stipend paid to the TBA is considered a “per diemâ€
to defray her personal expenses for the trip to the health facility. The amount of
the stipend was chosen because the volunteer community health workers (CHWs) in the
target area of the non-governmental organization that supports this project are paid
a “per diem†of KSH 200 per day of work. The number of TBAs participating in the referral
program was 38.
The number of births from the Kisiiki area, which was estimated by the number of new
PNC patients, was compared to the rural control facilities in the Yatta sub-county.
Data from the Ministry of Public Health and Sanitation (MOPHS) facilities’ records
was used to determine PNC visits and pregnant women who delivered at a rural health
facility with or without a TBA.
The control facilities consisted of 28 rural dispensaries and health centers within
the Yatta sub-county that did not participate in the TBA referral program. The two
intervention facilities for which results were reported in the previous study 16] were excluded from this analysis because we did not have data from them for each
time interval. The two urban hospitals were also excluded, since the focus of this
study is on rural health facilities. The number of first visits of PNC patients was
used to estimate the number of pregnancies because the actual number of pregnancies
cannot be determined from the available government statistics. Over 90Â % of pregnant
women are seen for at least one PNC visit, so the number of first PNC patient visits
approximates the number of viable pregnancies. Informed consent was obtained from
the participating TBAs.
The pre-intervention data was collected from July 2011 through February 2013. In the
control facilities, the number of new PNC patients and the number of deliveries were
2112 and 218 respectively. The intervention facility (Kisiiki) recorded 334 new PNC
patients and 12 deliveries during the same time period.
These pre-intervention figures were compared with post intervention data collected
from March 1, 2013 through May 31, 2013 (prior to enactment of the policy where maternity
services are free at health care facilities) and from June 1, 2013 through September
30, 2013 (after the enactment of the policy). Therefore, the post intervention period
was divided into subintervals, one before the free maternity care policy was enacted
and one after the policy went into effect. These periods were analyzed separately
to assess changes in SBA-attended birth rates related to the effect of the change
in the government policy.
Ethics
The University of New Mexico Human Research Protections Office and the Kenyatta National
Hospital-University of Nairobi Ethics and Research Committee in Kenya approved the
study.
Statistics
The SBA rate was calculated as a percentage of the prenatal care (PNC) patients who
delivered at the facility, with the denominator being the number of first visit prenatal
care patients and the numerator being the number of deliveries at the facility. The
total number of the PNC patients and deliveries at all non-intervention rural facilities
were used for the control percentages. Chi-square tests were used to test differences
in percent of SBA (facility) deliveries between intervention and control groups for
individual time periods: baseline rates (pre-intervention), for the rates during the
entire 7-month intervention period, and for the rates during the two parts of the
intervention period (before and after the implementation of the free maternity care
policy). Differential changes over time between the intervention and control groups
were tested using binomial regression with a term for the interaction between group
and time. Statistical analyses were performed using Stata version 13 (StataCorp LP,
College Station, TX, USA).