- Email: [email protected]
The relationship between women's experiences of mistreatment at facilities during childbirth, types of support received and person providing the support in Lucknow, India
Midwifery 40 (2016) 114–123
Contents lists available at ScienceDirect
Midwifery journal homepage: www.elsevier.com/midw
The relationship between women's experiences of mistreatment at facilities during childbirth, types of support received and person providing the support in Lucknow, India Nadia Diamond-Smith, PhD, MSc (Assistant Professor)a,n, May Sudhinaraset, PhD (Assistant Professor)a, Jason Melo, BS (Research Assistant)a, Nirmala Murthy, MS, DSc (advisor, FRHS)b a b
University of California, San Francisco, 550 16th Street, 3rd Floor, Box 1224, San Francisco, CA 94158, United States Foundation for Research in Health Systems, G-1, Brigade Bussiness Suites, 44, T.Mariappa Road, 100 feet Road, Jayanagar 2nd Block, Bangalore 560011, India
art ic l e i nf o
a b s t r a c t
Article history: Received 2 February 2016 Received in revised form 14 June 2016 Accepted 19 June 2016
Background: a growing body of literature has highlighted the prevalence of mistreatment that women experience around the globe during childbirth, including verbal and physical abuse, neglect, lack of support, and disrespect. Much of this has been qualitative. Research around the world suggests that support during childbirth can improve health outcomes and behaviours, and improve experiences. Support can be instrumental, informational, or emotional, and can be provided by a variety of people including family (husbands, mothers) or health providers of various professional levels. This study explores women's reported experiences of mistreatment during childbirth quantitatively, and how these varied by specific types of support available and provided by specific individuals. Methods: participants were women age 16–30 who had delivered infants in a health facility in the previous five years and were living in slums of Lucknow India. Data were collected on their experiences of mistreatment, the types of support they received, and who provided that support. Results: women who reported lack of support were more likely to report mistreatment. Lack of support in regards to discussions with providers and provider information were most strongly associated with a higher mistreatment score. Women who received any type of support from their husband or a health worker were significantly more likely to report lower mistreatment scores. Receiving informational support from a mother/mother-in-law or emotional support from a health worker was also associated with lower mistreatment scores. However, receiving emotional support from a friend/neighbour/other family member was associated with a higher mistreatment score. Conclusions: women rely on different people to provide different types of support during childbirth in this setting. Some of these individuals provide specific types of support that ultimately improve a woman's overall experience of her childbirth. Interventions aiming to reduce mistreatment to women during childbirth should consider the important role of increasing support for women, and who might be the most appropriate person to provide the most essential types of support through this process. & 2016 Published by Elsevier Ltd.
Keywords: Childbirth Mistreatment Support Abuse South Asia
Introduction There is a growing body of evidence on the mistreatment that some women experience during childbirth, including a recent systematic review on the topic (Bohren et al., 2015). Mistreatment can encompass a number of factors, including verbal and physical n
Corresponding author. E-mail addresses: [email protected] (N. Diamond-Smith), [email protected] (M. Sudhinaraset), [email protected] (J. Melo), [email protected] (N. Murthy). http://dx.doi.org/10.1016/j.midw.2016.06.014 0266-6138/& 2016 Published by Elsevier Ltd.
abuse, disrespect, and neglect of various forms (Vogel et al., 2015). Most of the current research on mistreatment is qualitative, and there are few quantitative studies, especially in low and middleincome countries. One mixed-methods study in India that included observations, interviews, surveys, mystery clients (simulated clients), and medical record reviews found evidence of many forms of mistreatment, including disrespect, abuse, and poor emotional support during childbirth (Hulton et al., 2007). There is increasing awareness of the importance of companionship and support for women during childbirth. For example, a systematic review of interventions that provided continuous
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
support by health workers or lay people to women during childbirth found that women who had support were more likely to have a spontaneous vaginal birth, and less likely to use anesthetics, have a infant with a low APGAR score, and report negative feelings about childbirth (Hodnett et al., 2012). They also had shorter labours than women who did not have one- on- one continuous support. One study in the US found that women who had a doula (a type of support in pregnancy) were 4 times less likely to have a low-birth weight infant and more likely to initiate breast feeding post partum (Gruber et al., 2013). Women who had doulas reported that they experienced less anxiety and had increased rates of breast feeding initiation than women who did not have doulas (Berghella et al., 2008; Hodnett et al., 2012). In addition to the more clinical outcomes described above, past research has also looked at women's experiences during childbirth. Continuous support from partners and providers were found to be important factors in improving psychosocial outcomes and coping with labour (Gibbins and Thomson, 2001). Social and emotional support can also be provided by family or friends. For example, a study in India found that women preferred the emotional support of a family member (generally a female relative, especially mothers-in-law or mothers) while delivering in facilities (Bhattacharyya et al., 2013). While most of the past research on the impact of support for women has been in the developed world, including all of the studies in the systematic review mentioned above, there is increasing awareness in the developing world, including India, on the importance of support during childbirth. For example, the government of the state of Tamil Nadu issued an order entitling women in government hospitals the right to have one companion in the delivery room. Qualitative interviews with women and support companions in Tamil Nadu, found that they felt that this would help protect them and their loved ones from abuse in health facilities; however, people were not aware of the government order allowing family members to be present (Sri, 2009). One intervention, called Yashoda, aimed at increasing support for women in India, placed a support worker or birth companion in facilities with high delivery volumes. The Yashoda intervention has been shown to improve exclusive breast feeding, family planning uptake, nutrition, immunisation, identification of danger signs and various aspects of postnatal care such as blood pressure and temperature checks of the infants (Varghese et al., 2014). Further research is needed to identify how support during childbirth is associated with mistreatment in order to improve women's experiences. Although the growing body of literature has noted high prevalence of mistreatment during childbirth in facilities, there has been no known quantitative research that measures the association between support (who was with a woman at the time of childbirth), what type of support that person provided, and the level of mistreatment a woman experienced. Few studies have attempted to disentangle what components of support are most helpful to women, and who specifically can provide these types of support most effectively. Qualitative interviews from women who experienced an infant death in New Delhi found that women felt that prohibiting partners from being allowed in the room for antenatal care or childbirth led to difficulty understanding what the doctor was telling them about their care and to problems convincing their partners to allow them to take the actions that the doctor instructed (Saikia et al., 2015). Another qualitative study from this study setting (Lucknow) included women who recently delivered, and described how husbands and mothers-in-law were yelled at and not allowed in the room with the woman during childbirth (Sudhinaraset et al., 2015). In the Indian context, female family members, usually mothers or mothers-in-law, accompany women to health facility during antenatal care and childbirth, but
115
are not allowed in the delivery room for reasons such as to avoid crowding. Increasingly, husbands are accompanying women to the facility for childbirth as they are expected to do tasks such as arranging transport, or getting blood and medicines. Their consent is also needed to undertake caesarean if the situation arises. For the last five years, a community level health worker (Accredited Social Health Activists or ASHAs), appointed by government, is expected to accompany rural women to health facilities for childbirth, for which she gets paid. Studies have not quantitatively assessed the type of support that different family members and close ties provide, and how it influences experiences of mistreatment. Social support is typically divided into three types of support: (1) emotional support, typically provided by a close intimate tie; (2) instrumental support, referring to aid or assistance of needs such as cooking, cleaning, paying bills; and (3) informational support (Berkman et al., 2000). Qualitative findings suggest that husbands and mother/mothersin-law may play different roles during the childbirth process. For example, husbands may be more likely to provide instrumental support, such as paying for bills and buying women needed medication, whereas mother/mothers-in-law provide emotional support during and after childbirth (Sudhinaraset et al., 2015). It appears that family members could potentially be important sources for women to get information and be able to act on that information, but there are other types of support, including emotional or instrumental (bringing the woman water, supplies) that might also be important. It is important to understand what types of support women are lacking and who currently provides specific type of support to women, given that women might be supported by their husbands, mothers or mothers-in-law, other family members, local health workers, or providers in the facility in this setting. Family members and health professionals can provide support before, during, and after childbirth, including both post partum in the facility and at home. Especially in India, where women are incentivized to stay in facilities for 48 hours after childbirth, their family members could provide important care and also help identify adverse outcomes if they arise in the postpartum period. Furthermore, it is important to understand if specific supporters are more protective of women experiencing mistreatment, not only adverse clinical outcomes.
Data and methods Data collection Data was collected from a cross-sectional sample in Lucknow, Uttar Pradesh, India in May, 2015 from a total of 759 young women living in economically disadvantaged (slum) areas. Eligibility criteria included having given birth in the previous five years and being aged 16–30 years at the time of the survey. This age range was chosen as we were particularly interested in whether younger women are more at risk of experiencing mistreatment and these are the prime childbearing years. An initial sampling frame of households in the slum areas was constructed, and from that, households were approached to determine if an eligible woman lived in that household. This analysis is limited to the 392 women from that sample who had delivered in a health facility (not at home). Household surveys were administered by four trained research assistants and covered a broad scope of topics including demographic characteristics, migration experiences, fertility, pregnancy, and childbirth experiences. Respondents were interviewed in their home or in another private setting selected by the respondent. Verbal informed consent was obtained from all study
116
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
participants due to the low literacy levels in these communities. All study documents were reviewed and approved by the Institutional Review Boards at the University of California, San Francisco, and Foundation for Research in Health Systems (FRHS) in India. Variable creation Mistreatment score We constructed an additive index score of mistreatment in health facilities using a series of questions from the survey regarding specific forms of mistreatment. The included questions asked women whether or not they experienced discrimination, physical or verbal abuse, threats to withhold treatment, lack of information, abandonment, their choice of position denied, requests for bribes, or unnecessary separation from the infant. These are subjective measures of a women's perception of her treatment. Answering yes to a question resulted in a score of 10 whereas answering no to a question resulted in a score of 0. If the respondent could not answer yes or no this was converted to a missing response. The total score any woman received was divided by the number of questions that they answered. Therefore, the mistreatment score of a woman who answered all the questions was comparable to that of a woman who was only able to answer some of the questions. Each woman received a score ranging from 0 to 10. The overall mean mistreatment score was 1.87 (SD 2.86). Support score Women were asked a series of questions about whether they received certain types of support and who provided that support, both during childbirth and in the postpartum period. This analysis focuses on the support provided during childbirth. Women were asked if someone did the following: brought them water or food; helped them talk to providers; provided them information about the infant; provided them support during labour; or provided encouragement during labour. While this might not be an exhaustive list of possible types of support, we felt it a representative sample of important areas of support, based on previous qualitative research on this subject in this setting. Women were then asked to indicate which people provided each type of support, and they could list more than one. Options included: mother-in-law, mother, husband, sister, friend/neighbour, ASHA/USHA, Anganwadi worker (community based health worker), Other health worker, and “other.” Specific “other” options commonly included brother, brother-in-law, father, father-in-law, and sister, which were coded as the “friend/neighbour” category, to expand this to “friend/neighbour/other family member”. Respondents also commonly mentioned doctors and nurses, who were recoded as “other health workers.” We also constructed a support score that combined the 5 components of support into a score (0–5). The first analysis presents women who reported that no one provided each type of support, and the second and third look at whether women did have support for each component, and who provided that support. Demographic and household variables All demographic variables were self-reported. Age was grouped into three categories: 16–19; 20–24; and 25–30. Women under 20 were separated, even though this makes the age ranges inconsistent, since we hypothesised that these women might be most at risk of mistreatment and lack of support. Women and their husband's education were grouped in three categories as well: no school/illiterate; primary; and secondary or more. Other variables included caste (scheduled caste, scheduled tribe, other backwards caste and no caste, as is the standard in this setting); religion
(Hindu and Muslim); number of living children; and whether the woman was currently employed. In addition, we constructed a wealth quartile variable in order to understand differences in support experiences in the context of comparative wealth among women. This variable was based on a series of survey questions regarding whether women possessed or had access to various appliances, technology, utilities, and vehicles. Examples included possessing a refrigerator, computer, running water, or a car respectively. Having answered yes to more questions would lead to a higher score; indicating a higher level of wealth. Scores were additive and the sample was divided equally into quartiles. These were based on the overall sample of 759 survey respondents and therefore, women who delivered in facilities (n¼ 392) were not equally distributed among the four quartiles. In ascending order of wealth, quartiles were labelled lowest quartile, second lowest quartile, second highest quartile, and highest quartile. Methods Three different regression analyses were run, controlling for the variables mentioned above. The first analysis examined the association between a woman reporting not having support and her reported mistreatment score (Table 4). The second analysis explored the association between a woman reporting that a specific person provided support and her reported mistreatment score (Table 5). The final analysis explored the relationship between different types of support provided by different individuals, and her reported mistreatment score. For this final analysis, we grouped the support into three categories: (1) Instrumental (bringing water or food); (2) Informational (providing information about infant or helping to talk to providers); and (3) Emotional (support or encouragement during labour) (Table 6).
Findings Basic demographics The quantitative survey included 392 women who reported that their most recent childbirth took place in a health facility in the last five years. Exactly half of the sample reported that this birth took place in a public sector primary health centre. Other reported childbirth sites were government hospitals (20.9%), private hospitals (20.9%), and private clinics (8.2%). A little more than half of women in this sample (54.7%) reported some form of mistreatment during a facility birth. The sampled women had a mean age of 25.3 years (SD ¼ 3.1 years) at the time of the survey, with 19 women (4.9%) less than 20 years of age, 134 (34.2%) aged 20–24 years, and 239 (60.9%) aged 25–30 years. Migrant women (women who had not lived in Lucknow their whole lives) were overrepresented, making up 92.9% of the sample. Approximately two-thirds of the women were Hindu (68.6%) with the remainder being Muslim. Most women belonged to a scheduled caste (33.4%), other backwards castes (40.6%) or scheduled tribe (11.7%) (the remaining had no caste). Almost the entire sample was currently married (99.5%) and reported living in households with a mean of 4.8 members (SD ¼1.7 members). While approximately a quarter of women reported achieving a secondary education or higher (25.8%), 148 women (37.8%) reported only having a primary education and 143 women (36.5%) reported having no education at all. Seventy-seven women (19.7%) fell into the poorest wealth quartile, followed by 88 (22.5%) in the lower middle quartile, 101 (25.8%) in the upper middle quartile, and 125 (32.0%) in the highest quartile.
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 1 Women who reported they had no one to provide the following types of care/ support at childbirth. N Brought them water/food Talk to providers Provide information about the infant Provide labour support Provide encouragement
Score of lack of support 0 1 2 3 4
%
1 84 63 50 38
0.3 21.4 16.1 12.8 9.7
216 115 49 11 1
55.1 29.3 12.5 2.8 0.3
117
(59.7%), followed by mothers (33.9%), sisters/brothers/fathers (including -in-laws) (28.8%) and mothers-in-laws (21.4%) (Table 3). Health workers or friends rarely brought water or food. Husbands were the most common person to help talk to providers (43.4%), followed by health workers (20.4%), mothers (29.1%), mothers-inlaw (15.8%) and other family members (11.2%). In terms of providing information about the infant, health workers were the most common person to do this (41.6%), followed by the woman's mother (34.9), and then her mother-in-law (19.4%). Support for labour and/or breathing was most often provided by health workers (54.3%), followed by mothers (30.4%) and mothers-in-law (21.2%). Finally, mothers, husbands, and health workers all provided encouragement or guidance in labour at about the same frequency (37.8%, 37.2% and 36.7%, respectively). Mothers-in-law also frequently provided support (29.8%), followed by friends (17.65%). Multivariate regressions
Experiences of support Support during childbirth Only one woman reported that she had no one to bring her water during childbirth (Table 1). Eighty-four women (21.4%) said they had no one to help them talk to providers, 63 (16.1%) said they had no one to provide information about the infant, 50 (12.8%) had no one to provide support in labour and 38 (9.7%) had no one to provide encouragement in labour. Overall, 55.1% (216) women reported someone provided all types of support, 29.3% had no one for 1 type of support, 12.5% for two types, 2.8% for three types and 0.3% for 1 type.
Lack of support during childbirth and mistreatment Having no one to talk to providers (2.192, p o0.01) or provide information about the infant (1.606, p o0.01) was associated with a higher mistreatment score, after controlling for the demographic factors (Table 4). Reporting more experiences of lack of support overall was also associated with reporting more mistreatment (0.907, p o0.01). Being from scheduled tribe or other backwards caste was consistently associated with higher mistreatment scores, compared to women from other castes. Additionally, being from the highest wealth quintile (compared to lowest) was associated with reporting higher mistreatment scores for all types of lack of support.
Person who provided support Most women, 77.8% (305) women reported that their husband provided them support for at least one of the indicators (Table 2). Most women, 80.4% (315), said that their mother-in-law or mother provided them support for at least 1 indicator, 59.9% (235) that their sister, friend or other, and about 52.8% (207) that a health worker such as an ASHA, anganwadi or other provided this support.
Person who provided support and mistreatment Reporting that her husband ( 1.384, po 0.01) or a health worker ( 0.879, p o0.01) was the person who provided at least one type of support during childbirth was associated with a lower mistreatment score (Table 5). Reporting that her mother or mother-in-law or a sister, friend or other person provided at least 1 type of support was not associated with the mistreatment score.
Type of support provided by each person Water and/or food was most often supplied by husbands Table 2 Women who indicated that they had support from the following people for at least 1 aspect of care/support during childbirth.
Husband Health worker (ASHA, anganwadi, other) Sister, friend, other Mother-in-law, mother
N
%
305 207 235 315
77.8 52.8 59.9 80.4
Types of support provided by specific people and mistreatment No types of support provided by a woman's husband were associated with the mistreatment score (Table 6). Having either their mother or mother-in-law provide informational support was significantly associated with the woman reporting a lower mistreatment score ( 0.835, p o0.05). Reporting that a health worker provided emotional support was associated with a lower reported mistreatment score ( 0.723, p o0.05). Different types of support from other family members or friends had different associations with mistreatment. An other family member or friend providing emotional support was associated with increased mistreatment scores (1.409, p o0.01), but one of these people providing
Table 3 Percent of women reporting types of support provided by each provider of support.
MIL Mother Husband Sister/Brother/Father (including in-laws) Friend All Health Workers
Instrumental
Informational
Emotional
Brought food or Water (%)
Helped Talk to providers (%)
Info about infant Support for labour, (%) breathing (%)
Encouragement or guidance (%)
21.4 33.9 59.7 28.8
15.8 29.1 43.4 11.2
19.4 34.9 7.1 7.4
21.2 30.4 0.8 2.8
29.8 37.8 37.2 8.4
7.9 2.6
5.6 20.4
8.2 41.6
9.4 54.3
17.6 36.7
118
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 4 Association between mistreatment score and lack of support during childbirth (Odds Ratio, SD). No one to pro- No one to help vide water talk to providers
No one to Provide information about the infant
No one to provide labour support
No one to provide encouragement
Score of lack of support during childbirth
0.857 (2.799)
2.192n (0.348)
1.606n (0.404)
0.248 (0.451)
0.237 (0.505)
0.907n (0.181)
0.788 (0.692)
0.643 (0.658)
0.561 (0.680)
0.779 (0.692)
0.769 (0.694)
0.629 (0.670)
25–30 years
0.256 (0.698)
0.190 (0.662)
0.177 (0.683)
0.237 (0.698)
0.229 (0.699)
0.193 (0.675)
Number of Living Children
0.111 (0.142)
0.127 (0.135)
0.111 (0.139)
0.104 (0.143)
0.107 (0.142)
0.111 (0.138)
0.245 (0.355)
0.392 (0.338)
0.207 (0.347)
0.228 (0.355)
0.249 (0.355)
0.185 (0.343)
0.00988 (0.452)
0.0263 (0.429)
0.0748 (0.442)
0.00843 (0.452)
0.00810 (0.452)
0.0922 (0.437)
0.0472 (0.366)
0.191 (0.347)
0.110 (0.359)
0.0296 (0.365)
0.0374 (0.365)
0.00344 (0.353)
Secondary or more
0.397 (0.412)
0.576 (0.393)
0.468 (0.404)
0.391 (0.412)
0.399 (0.412)
0.469 (0.399)
Currently works outside the home
0.162
0.0715
0.161
0.171
0.172
0.0469
(0.318)
(0.304)
(0.311)
(0.317)
(0.318)
(0.308)
0.185
0.0698
0.0744
0.19
0.191
0.111
(0.338)
(0.321)
(0.331)
(0.338)
(0.338)
(0.326)
0.286 (0.487)
0.366 (0.463)
0.202 (0.477)
0.284 (0.487)
0.314 (0.490)
0.184 (0.471)
Scheduled Tribe
1.769n (0.590)
2.017n (0.561)
1.276† (0.590)
1.703n (0.602)
1.818n (0.598)
1.094§ (0.586)
Other backwards caste
1.199† (0.474)
1.258n (0.450)
0.925† (0.468)
1.186† (0.473)
1.211† (0.475)
0.978† (0.460)
Migrant (compared to nonmigrant)
0.0223
0.343
0.211
0.0213
0.0212
0.262
(0.566)
(0.541)
(0.557)
(0.566)
(0.566)
(0.550)
0.387 (0.446)
0.220 (0.423)
0.334 (0.436)
0.360 (0.446)
0.400 (0.447)
0.110 (0.434)
0.534 (0.442)
0.557 (0.420)
0.516 (0.433)
0.513 (0.444)
0.548 (0.443)
0.388 (0.429)
Mistreatment Score
Age group (compared to 16–19 years) 20–24 years
Woman's Education level (compared to no education/ illiterate) Primary
Secondary or more
Man's Education level (compared to no education/ illiterate) Primary
Religion (Muslim compared to Hindu)
Caste (compared to no caste) Scheduled Caste
Wealth quartile (lowest quartile) Second-lowest quartile
Second highest quartile
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
119
Table 4 (continued ) No one to pro- No one to help vide water talk to providers
No one to Provide information about the infant
No one to provide labour support
No one to provide encouragement
Score of lack of support during childbirth
Highest quartile
1.753n (0.462)
1.316n (0.444)
1.513n (0.456)
1.726n (0.464)
1.766n (0.462)
1.225n (0.459)
Constant
0.464 (1.113)
0.547 (1.057)
0.726 (1.092)
0.467 (1.113)
0.48 (1.114)
0.654 (1.077)
Observations R-squared
380 0.109
380 0.197
380 0.146
380 0.11
380 0.11
380 0.167
n
† §
po 0.01. p o 0.05. p o 0.1.
informational support was associated with lower reported mistreatment scores ( 0.855, po 0.05). None of the other types of support were associated with mistreatment scores for any people providing the support. Interestingly, instrumental support provided by any of the sources of support was not associated with the mistreatment score.
Discussion This is one of the first studies to explore in detail both who provides support to women and what type of support is provided by different people who accompany women during childbirth at facilities. Taking this one step further, we were also able to then see how lack of support overall and specific types of support being provided by specific sources of support (people), were associated with women's experiences of mistreatment. The majority of women in the sample had some form of support for all measures. The most common type of support that women were lacking during childbirth was having someone to help them talk to providers (21%). This is relevant in respect to past qualitative research where women discussed not being able to understand what the providers wanted them to do and how their partners being refused access made it harder to act on doctors' advice (Saikia et al., 2015). About 1 in 10 women reported not having anyone to provide labour support or encouragement. While this is a minority of the women, it still means that a substantial number of women are delivering without the emotional support that past literature has suggested is associated with improved physical and mental health outcomes. Interestingly, mothers-in-laws/mothers (80%) and husbands (79%), were the most common people to provide support, and health workers only provided support to about half (53%) of women. It appears that despite restrictions on physical accessibility to the woman during childbirth, family members are still the most frequent provider of support to women. It is also alarming that women do not report that health care providers are providing informational support to them in this period given that providers should be trained in counselling women on health care processes. It is possible that providers are simply not giving women information about what is happening to them during their labour and delivery, perhaps due to limited time, skill or willingness. However, it is also possible that providers are providing information to women in a way that women are not able to absorb or understand, perhaps due to the use of highly technical terms, not enough time for women to ask questions or providers to make sure women understand, or other barriers. A significant literature has described system-level challenges in supporting healthcare
providers to deliver the best care (Jewkes et al., 1998; Kruk et al., 2010; Freedman and Kruk, 2014). The failure of health systems for women during childbirth occurs at every level – from the facility to national and policy level (Freedman and Kruk, 2014). Providers may be mistreating women because they themselves do not have the time, resources, or system support needed to effectively perform their work duties. Jewkes et al. (1998) provides an overview of nurse–woman relationships in South African obstetric care, and finds that humiliation and physical abuse of women was quite common in this setting (Jewkes et al., 1998). Explanations for these behaviours included system level challenges such as nurses struggling to assert their power among lower class women, lack of accountability of services and lack of quality assurance by facility managers and higher-level providers, and overall culture of abuse in nursing care. Improving quality assurance and accountability at the highest levels may improve normative behaviours at the facility level to improve the care of women during labour and delivery (Freedman and Kruk, 2014). We find that lower caste women are more likely to report mistreatment (discussed in more detail in a forthcoming paper focusing on this from the same data set), suggesting that power dynamics in terms of the interaction between women and health facility staff may be at play in this setting as well (Sudhinaraset et al., 2016). Even though only about half of women reported having support from a health worker, it appears that the support of these people is important for reducing the experience of mistreatment. When health care providers are playing a positive role in women's childbirth experience, such as by providing information or encouragement, this improves women's overall view of her childbirth experience. Health care providers are already in these facilities, so engaging them more in providing types of support that women are currently lacking could have a big impact on overall experiences. However, these health care providers are often overworked and underpaid, especially in public facilities, therefore, it could prove challenging to encourage them to show more care to women at the time of childbirth without providing them additional support (World Health Organization, 2006). A landscape analysis of factors contributing to abuse and disrespect in childbirth highlighted factors at various levels, including individual/community; national laws and policies, human rights and ethics; governance and leadership; service delivery; and providers (Bowser and Hill, 2010). They discussed in detail how lack of standards, poor accountability, provider demoralisation, shortages of human resources and lack of professional development opportunities and low provider status and respect all contribute to abuse and disrespect. Lack of informational support (no one to help talk to providers or provide information about the infant) was most strongly
120
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 5 Association between mistreatment score and person providing support during childbirth (Odds Ratio, SD). Husband Health worker
Mistreatment score
Age group (compared to 16–19 years) 20–24 years
1.384 (0.344)
n
0.879 (0.300)
Sister/ Friend/Other
n
Mother/ Mother-inlaw
0.472 (0.298)
0.361 (0.375)
Table 5 (continued ) Husband Health worker
Sister/ Friend/Other
Mother/ Mother-inlaw
Second highest quartile
0.606 (0.433)
0.602 (0.438)
0.538 (0.441)
0.537 (0.442)
Highest quartile
1.652n (0.452)
1.747n (0.456)
1.802n (0.461)
1.722n (0.462)
0.732 (0.677)
0.846 (0.684)
0.841 (0.691)
0.780 (0.691)
Constant
1.932§ (1.149)
0.948 (1.113)
0.276 (1.115)
0.796 (1.167)
25–30 years
0.208 (0.683)
0.258 (0.689)
0.230 (0.695)
0.226 (0.697)
Observations R-squared
380 0.147
380 0.130
380 0.115
380 0.112
Number of Living Children
0.107
0.0930
0.108
0.111
n
§
(0.139)
Woman's Education level (compared to no education/illiterate) Primary
(0.142)
0.241 (0.354)
0.0594 0.114 (0.442) (0.448)
0.00844 (0.450)
0.0357 (0.452)
0.00742 (0.357)
0.181 (0.364)
0.00408 (0.365)
0.0320 (0.365)
Secondary or more
0.457 (0.404)
0.379 (0.408)
0.473 (0.414)
0.406 (0.412)
Currently works outside the home
0.00270
0.112
0.186
0.155
(0.313)
(0.314)
(0.317)
(0.317)
0.203
0.156
0.218
0.197
(0.330)
(0.333)
(0.337)
(0.337)
0.235 (0.477)
0.399 (0.483)
0.318 (0.486)
0.307 (0.487)
Scheduled Tribe
1.708n (0.577)
1.947n (0.586)
1.863n (0.590)
1.665n (0.599)
Other backwards caste
1.073† (0.464)
1.406n (0.474)
1.215† (0.472)
1.184† (0.473)
Migrant (compared to non-migrant)
0.102
0.167
0.0243
0.0458
(0.555)
(0.563)
(0.565)
(0.566)
0.376 (0.435)
0.307 (0.441)
0.320 (0.445)
0.354 (0.445)
Religion (Muslim compared to Hindu)
Caste (compared to no caste) Scheduled Caste
Wealth quartile (lowest quartile) Second-lowest quartile
p o 0.01. po 0.05. p o0.1.
(0.142)
0.284 (0.355)
Man's Education level (compared to no education/illiterate) Primary
0.265 (0.347)
(0.141)
0.443 (0.357)
Secondary or more
†
associated with reporting a higher mistreatment score. On the other hand, instrumental and emotional support were not associated with higher levels of mistreatment (except increased mistreatment reported when emotional support was provided by other family or friends). This suggests that women who understand what is happening during childbirth, including decisions made about their care, have better experiences during childbirth. This finding supports broader calls for person-centred care approaches, which highlights the need to include women in the decision-making process and inform women on all aspects of care. Husbands were the person most frequently reported as providing instrumental support (getting water), although other family members, including mothers and mothers-in-law also provided this type of support. The importance of husbands’ presence and support in improving women's experiences in childbirth has been shown elsewhere. For example, past qualitative research in India also found that women described their husbands as providing instrumental support, such as helping arrange transportation and buy supplies when needed for childbirth (Sudhinaraset et al., 2015). Research in Nepal found that women reported higher scores on a measure of agency during labour and delivery when their husbands were with them than when female friends were with them (although this was also associated with higher scores on the agency scale overall) (Sapkota et al., 2012). In our study, husbands were the most frequent person recorded as having helped the woman talk to providers, although mothers also provided this type of support. The provision of this type of support could explain previous findings about increased feelings of agency with husband's presence. Informational support (providing information about the infant) was much more frequently provided by health workers and mothers/ mothers-in-law and rarely by husbands. This is to be expected since husbands might be much less likely to be as deeply involved in infant-care. Women reported that health providers were common sources of support and encouragement during labour, as were mothers and mothers-in-law. Husbands provided encouragement, however, they were not frequently reported as providing support during labour. This is intuitive since the support question was about specific types of support (with breathing or techniques in labour), and husbands might have less knowledge about these very female-specific needs, although male partners in other settings have been able to provide women this type of support. Mothers and mothers-in-law provided the highest level of
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 6 Association between types of support provided and mistreatment score, by person providing the support (Odds Ratio, SD).
Instrumental Support
Emotional Support
Informational Support
Husband Mother/Mo- Health ther-in-law worker
Other Family member /Friend
0.422 (0.309)
0.344 (0.305)
0.104 (0.311)
1.128 (0.901) n
Table 6 (continued ) Husband Mother/Mo- Health ther-in-law worker
Other Family member /Friend
0.443 (0.446)
0.213 (0.448)
0.298 (0.439)
0.400 (0.435)
Second highest quartile
0.569 (0.440)
0.518 (0.441)
0.664 (0.437)
0.764§ (0.433)
Highest quartile
1.654† (0.462)
1.626† (0.463)
1.632† (0.456)
1.698† (0.450)
Wealth quartile (lowest quartile) Second-lowest quartile
†
0.391 (0.312)
0.347 (0.331)
0.723 (0.352)
1.490 (0.345)
0.407 (0.322)
0.835n (0.348)
0.427 (0.345)
0.855n (0.355)
0.744 (0.690)
0.920 (0.682)
0.798 (0.674)
Constant
1.406 (1.175)
0.805 (1.154)
1.175 (1.113)
0.252 (1.086)
Observations R-squared
380 0.128
380 0.124
380 0.142
380 0.162
Age group (compared to 16–19 years) 20–24 years 0.765 (0.692)
25–30 years
0.342 (0.695)
0.207 (0.696)
0.470 (0.691)
0.109 (0.682)
Number of Living Children
0.124
0.0841
0.115
0.0696
n
0.0559 (0.445)
0.0384 (0.441)
0.0121 (0.363)
0.0375 (0.365)
0.143 (0.362)
0.121 (0.357)
Secondary or more
0.437 (0.410)
0.434 (0.411)
0.444 (0.406)
0.537 (0.405)
Currently works outside the home
0.0598
0.179
0.0739
0.183
(0.317)
(0.316)
(0.314)
(0.309)
0.185
0.203
0.119
0.0495
(0.336)
(0.336)
(0.332)
(0.330)
0.149 (0.492)
0.302 (0.485)
0.310 (0.480)
0.332 (0.478)
Scheduled Tribe
1.700† (0.601)
1.547n (0.602)
1.873† (0.583)
1.782† (0.575)
Other backwards caste
1.089n (0.473)
1.137n (0.473)
1.172n (0.468)
0.964n (0.467)
Migrant (compared to non-migrant)
0.191
0.0849
0.196
0.121
(0.574)
(0.564)
(0.561)
(0.555)
Caste (compared to no caste) Scheduled Caste
p o 0.05. po 0.01. p o 0.1.
(0.139)
0.0431 0.0343 (0.449) (0.450)
Religion (Muslim compared to Hindu)
0.250 (0.354)
(0.141)
0.262 (0.348)
Woman's Education level (compared to no education/illiterate) Primary
0.248 (0.353)
(0.142)
0.368 (0.352)
Secondary or more
† §
(0.141)
Woman's Education level (compared to no education/illiterate) Primary
121
support across the different categories (instrumental, informational and emotional). A number of studies in South Asia have highlighted the gate-keeper role of mothers-in-laws in reproductive health care decision-making and practices, including antenatal care use, place of childbirth, family planning use, abortions, and postnatal practices including breast feeding (Barua and Kurz, 2001; Ganatra and Hirve, 2002; Masvie, 2006; Char et al., 2010; Diamond-Smith et al., 2012; Sudhinaraset et al., 2015). However, few studies have documented the support that mothers and mothers-in-law are providing to women during childbirth. One study on Kangaroo Mother Care (KMC) in Chandigarh, India found that mothers-in-law provided KMC for over half of women (Parmar et al., 2009). It is clear that a better understanding is necessary of what role mothers and mothers-in-law play at the facility in providing care and support to women, especially for planning interventions that might want to target these already existing forms of support to engage them in providing additional support for women. Women who received support from husbands and health workers had lower mistreatment scores, while there was no association between mistreatment and receiving support from mothers/mothers-in-law and other family members/friends. Given the patriarchal nature of Indian society, it seems plausible that health providers treat women with more respect when their husbands are actively engaged in their childbirth and that the wishes and desires of women are more often fulfilled if their husband is supporting them and advocating for them. This suggests that altering policies to encourage (or allow) husbands/ partners to be able to accompany women could reduce experiences of mistreatment. It is possible that providers, who are often female (in Indian most providers for women during childbirth at all levels are female), might also be more afraid of abusing, disrespecting, or demanding bribes of women (components of the mistreatment score) when their husbands are present. Although husbands frequently provided instrumental support, this was not associated with the woman reporting a lower mistreatment score. Mothers or mothers-in-law providing informational support (helping communicate with providers or providing information about infant care) was significantly associated with
122
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
lower mistreatment scores, and these individuals were frequently reported as providing these types of support although they were not the most common person for either category. The only person whose provision of emotional support was significantly associated with lower mistreatment scores was a health worker. One study (in the US) looking at both student midwives and laywomen as support for women during childbirth found that women were satisfied with both types of people playing supportive roles (Hemminki et al., 1990). Women who did not have their husband with them were especially positive about having health workers as support. However, the student midwives did not see much value in playing this role for women and were generally negative about the trial. In our study, contrary to the situation with health providers, having emotional support from other family members or friends was associated with higher mistreatment scores. This is in contrast to past literature, which found that support from a friend or close family member was associated with more satisfaction, which, although a different measure, is associated with mistreatment (Hodnett et al., 2012). It is possible that emotional support is needed from other family or friends when the childbirth experience is especially bad. In other words, it is possible that the relationship goes the other way, where the experience of mistreatment leads women to call upon their family or friends to provide them emotional support. There are several limitations to this study. First, data was only collected from women ages 16-30 living in slum areas of Lucknow, and therefore is unlikely to be generalisable to richer women in Lucknow or to women in other parts of Uttar Pradesh or India. Furthermore, migrant women were overrepresented in the survey, therefore, these findings might not be reflective of non-migrant women. All data was collected retrospectively, as women had delivered in the last 5 years, and therefore might be subject to recall bias. Our measure of mistreatment is based on women's selfreports of their own experiences of mistreatment, which is a subjective measure and thus the same experience might be interpreted differently by different types of women. We attempt to control for this by including variables in the analysis such as education and wealth, as past literature has suggested that richer and more educated women report higher levels of mistreatment. We are aware that women may interpret the meaning of the wording about “information” or “respect” differently, however, our interest is a woman's experience (perception, internalisation), rather than what actually happened. Future research could use mixed methods combining these measures with observations, in order to triangulate findings on perceived and actual experiences. Also related to women's perceptions of mistreatment, it is possible that women that had more adverse post partum outcomes might have had more unfavourable opinions of their care looking back. We were unable to analyse any associations between mistreatment and adverse maternal or child health outcomes, due to the small sample size and relative rarity of these events. We recommend that future research explore how mistreatment is associated with outcomes such as complications in labour and use of postnatal services in this population. We also suggest that research in westernized countries focused more on women's experiences of childbirth, in addition to clinical and behavioural outcomes. The mounting body of evidence on mistreatment during childbirth, which this research adds to, needs to be funneled into interventions aimed to reduce mistreatment and improve women's childbirth experiences. Support for women is especially important because this is a component of mistreatment that is associated with improved outcomes and behaviours. Very little attention has been paid to how support and mistreatment are interlinked in global contexts, and especially in low and middleincome countries, despite the attention on mistreatment broadly. This research provides an important contribution in that it not
only looks at the relationship between the absence of specific types of support (instrumental, informational and emotional) and mistreatment, but it takes this one step further by looking at what specific people provide support, and when the support of specific people for specific domains actually are most able to deter mistreatment. We recommend that interventions be developed that turn these findings into action by engaging different family members and health workers to provide specific types of support to women during childbirth. Utilising family members and health workers to provide support provides a low-hanging fruit, since these individuals are already often with women at childbirth, and thus need only be incentivized and engaged to uphold the professional standards of practice, quality monitoring and regulation. We also recommend that institutional policy regarding support people for women during childbirth be expanded to acknowledge the importance of and increase support to women during childbirth, and highlight the necessity of allowing multiple people to provide different forms of support to women.
Acknowledgements The authors would like to acknowledge the Bill and Melinda Gates Foundation (Grant ID: OPP1084330) for the funding for this project. We would also like to acknowledge Kanksha Singh, Emily Treleaven, and Rajni Wadwa.
References Barua, A., Kurz, K., 2001. Reproductive health-seeking by married adolescent girls in Maharashtra, India. Reproductive Health Matters 9, 53–62. http://dx.doi.org/ 10.1016/S0968-8080(01)90008-4. Berghella, V., Baxter, J.K., Chauhan, S.P., 2008. Evidence-based labor and delivery management. American Journal of Obstetrics and Gynecology 199, 445–454. http://dx.doi.org/10.1016/j.ajog.2008.06.093. Berkman, L.F., Glass, T., Brissette, I., Seeman, T.E., 2000. From social integration to health: Durkheim in the new millennium. Social Science & Medicine 51, 843–857. http://dx.doi.org/10.1016/S0277-9536(00)00065-4. Bhattacharyya, S., Srivastava, A., Iqbal Avan, B., 2013. Delivery should happen soon and my pain will be reduced: understanding women's perception of good delivery care in India. Global Health Action 6. http://dx.doi.org/10.3402/gha. v6i0.22635. Bohren, M.A., Vogel, J.P., Hunter, E.C., Lutsiv, O., Makh, S.K., Souza, J.P., et al., 2015. The mistreatment of women during childbirth in health facilities globally: a mixed-methods systematic review. PLoS Medicine 12 (6), e1001847, 1-32, http://dx.doi.org/10.1371/journal.pmed.1001847. Bowser, D., Hill, K., 2010. Exploring Evidence for Disrespect and Abuse in FacilityBased Childbirth: Report of a Landscape Analysis. USAID. Char, A., Saavala, M., Kulmala, T., 2010. Influence of mothers-in-law on young couples’ family planning decisions in rural India. Reproductive Health Matters 18, 154–162. http://dx.doi.org/10.1016/S0968-8080(10)35497-8. Diamond-Smith, N., Campbell, M., Madan, S., 2012. Misinformation and fear of sideeffects of family planning. Culture Health & Sexuality 14, 421–433. http://dx. doi.org/10.1080/13691058.2012.664659. Freedman, L.P., Kruk, M.E., 2014. Disrespect and abuse of women in childbirth: challenging the global quality and accountability agendas. Lancet 384, e42–e44. http://dx.doi.org/10.1016/S0140-6736(14)60859-X. Ganatra, B., Hirve, S., 2002. Induced abortions among adolescent women in rural Maharashtra, India. Reproductive Health Matters 10, 76–85. http://dx.doi.org/ 10.1016/S0968-8080(02)00016-2. Gibbins, J., Thomson, A.M., 2001. Women's expectations and experiences of childbirth. Midwifery 17, 302–313. http://dx.doi.org/10.1054/midw.2001.0263. Gruber, K.J., Cupito, S.H., Dobson, C.F., 2013. Impact of Doulas on healthy birth outcomes. The Journal of Perinatal Education 22, 49–56. Hemminki, E., Virta, A.-L., Koponen, P., Malim, M., Kojo-Austin, H., Tuimala, R., 1990. A trial on continuous human support during labor: feasibility, interventions and mothers' satisfaction. Journal of Psychosomatic Obstetrics & Gynecology 11, 239–250. http://dx.doi.org/10.3109/01674829009084420. Hodnett, E., Gates, S., Hofmeyr, G., Sakala, C., 2012. Continuous support for women during childbirth. Cochrane Database of Systematic Reviews, 10. Hulton, L.A., Matthews, Z., Stones, R.W., 2007. Applying a framework for assessing the quality of maternal health services in urban India. Social Science & Medicine 64, 2083–2095. http://dx.doi.org/10.1016/j.socscimed.2007.01.019. Jewkes, R., Abrahams, N., Mvo, Z., 1998. Why do nurses abuse patients? Reflections from South African obstetric services. Social Science & Medicine 1982,
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
1781–1795. Kruk, M.E., Rockers, P.C., Mbaruku, G., Paczkowski, M.M., Galea, S., 2010. Community and health system factors associated with facility delivery in rural Tanzania: a multilevel analysis. Health Policy 97, 209–216. http://dx.doi.org/10.1016/j. healthpol.2010.05.002. Masvie, H., 2006. The role of Tamang mothers-in-law in promoting breast feeding in Makwanpur District, Nepal. Midwifery 22, 23–31. http://dx.doi.org/10.1016/j. midw.2005.02.003. Parmar, V.R., Kumar, A., Kaur, R., et al., 2009. Experience with Kangaroo mother care in a neonatal intensive care unit (NICU) in Chandigarh, India. Indian Journal of Pediatrics 76, 25–28. http://dx.doi.org/10.1007/s12098-009-0024-2. Saikia, N., Sehgal, S., Kumar Bora, J., Diamond-Smith, N., 2015. Why are Neonatal Deaths Higher among Urban Poor? Qualitative Insights from Metropolitan Delhi. Rev. Sapkota, S., Kobayashi, T., Kakehashi, M., Baral, G., Yoshida, I., 2012. In the Nepalese context, can a husband's attendance during childbirth help his wife feel more in control of labour? BMC Pregnancy and Childbirth 12, 49. http://dx.doi.org/ 10.1186/1471-2393-12-49. Sri, S.B., 2009. Translating medical evidence into practice: working with
123
communities and providers to promote active management of the third stage of labour. Health Popul. Innov. Fellowsh. Programme Working paper 11. Sudhinaraset, M., Beyler, N., Barge, S., Nadia, 2015. Decision-making for delivery location and quality of care among Slum-Dwellers: a qualitative study in Uttar Pradesh, India. Rev. BMC Pregnancy Childbirth. Sudhinaraset, M., Treleavan, E., Melo, J., Singh, K., Diamond-Smith, N., n.d. Women's Status and Experiences of Mistreatment during Childbirth in Uttar Pradesh: A Mixed Methods Study using Cultural Health Capital Theory. Rev., 2016. Varghese, B., Roy, R., Saha, S., Roalkvam, S., 2014. Fostering maternal and newborn care in India the Yashoda Way: does this improve maternal and newborn care practices during institutional delivery? PLoS One 9. http://dx.doi.org/10.1371/ journal.pone.0084145. Vogel, J., Bohren, M., Tunçalp, Ò, Oladapo, O., Gülmezoglu, A., 2015. Promoting respect and preventing mistreatment during childbirth. BJOG: An International Journal of Obstetrics and Gynaecology . http://dx.doi.org/10.1111/ 1471-0528.13750, n/a–n/a. World Health Organization, 2006. Working Together for Health: World Health Report 2006. The World Health Organization, Geneva.
Contents lists available at ScienceDirect
Midwifery journal homepage: www.elsevier.com/midw
The relationship between women's experiences of mistreatment at facilities during childbirth, types of support received and person providing the support in Lucknow, India Nadia Diamond-Smith, PhD, MSc (Assistant Professor)a,n, May Sudhinaraset, PhD (Assistant Professor)a, Jason Melo, BS (Research Assistant)a, Nirmala Murthy, MS, DSc (advisor, FRHS)b a b
University of California, San Francisco, 550 16th Street, 3rd Floor, Box 1224, San Francisco, CA 94158, United States Foundation for Research in Health Systems, G-1, Brigade Bussiness Suites, 44, T.Mariappa Road, 100 feet Road, Jayanagar 2nd Block, Bangalore 560011, India
art ic l e i nf o
a b s t r a c t
Article history: Received 2 February 2016 Received in revised form 14 June 2016 Accepted 19 June 2016
Background: a growing body of literature has highlighted the prevalence of mistreatment that women experience around the globe during childbirth, including verbal and physical abuse, neglect, lack of support, and disrespect. Much of this has been qualitative. Research around the world suggests that support during childbirth can improve health outcomes and behaviours, and improve experiences. Support can be instrumental, informational, or emotional, and can be provided by a variety of people including family (husbands, mothers) or health providers of various professional levels. This study explores women's reported experiences of mistreatment during childbirth quantitatively, and how these varied by specific types of support available and provided by specific individuals. Methods: participants were women age 16–30 who had delivered infants in a health facility in the previous five years and were living in slums of Lucknow India. Data were collected on their experiences of mistreatment, the types of support they received, and who provided that support. Results: women who reported lack of support were more likely to report mistreatment. Lack of support in regards to discussions with providers and provider information were most strongly associated with a higher mistreatment score. Women who received any type of support from their husband or a health worker were significantly more likely to report lower mistreatment scores. Receiving informational support from a mother/mother-in-law or emotional support from a health worker was also associated with lower mistreatment scores. However, receiving emotional support from a friend/neighbour/other family member was associated with a higher mistreatment score. Conclusions: women rely on different people to provide different types of support during childbirth in this setting. Some of these individuals provide specific types of support that ultimately improve a woman's overall experience of her childbirth. Interventions aiming to reduce mistreatment to women during childbirth should consider the important role of increasing support for women, and who might be the most appropriate person to provide the most essential types of support through this process. & 2016 Published by Elsevier Ltd.
Keywords: Childbirth Mistreatment Support Abuse South Asia
Introduction There is a growing body of evidence on the mistreatment that some women experience during childbirth, including a recent systematic review on the topic (Bohren et al., 2015). Mistreatment can encompass a number of factors, including verbal and physical n
Corresponding author. E-mail addresses: [email protected] (N. Diamond-Smith), [email protected] (M. Sudhinaraset), [email protected] (J. Melo), [email protected] (N. Murthy). http://dx.doi.org/10.1016/j.midw.2016.06.014 0266-6138/& 2016 Published by Elsevier Ltd.
abuse, disrespect, and neglect of various forms (Vogel et al., 2015). Most of the current research on mistreatment is qualitative, and there are few quantitative studies, especially in low and middleincome countries. One mixed-methods study in India that included observations, interviews, surveys, mystery clients (simulated clients), and medical record reviews found evidence of many forms of mistreatment, including disrespect, abuse, and poor emotional support during childbirth (Hulton et al., 2007). There is increasing awareness of the importance of companionship and support for women during childbirth. For example, a systematic review of interventions that provided continuous
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
support by health workers or lay people to women during childbirth found that women who had support were more likely to have a spontaneous vaginal birth, and less likely to use anesthetics, have a infant with a low APGAR score, and report negative feelings about childbirth (Hodnett et al., 2012). They also had shorter labours than women who did not have one- on- one continuous support. One study in the US found that women who had a doula (a type of support in pregnancy) were 4 times less likely to have a low-birth weight infant and more likely to initiate breast feeding post partum (Gruber et al., 2013). Women who had doulas reported that they experienced less anxiety and had increased rates of breast feeding initiation than women who did not have doulas (Berghella et al., 2008; Hodnett et al., 2012). In addition to the more clinical outcomes described above, past research has also looked at women's experiences during childbirth. Continuous support from partners and providers were found to be important factors in improving psychosocial outcomes and coping with labour (Gibbins and Thomson, 2001). Social and emotional support can also be provided by family or friends. For example, a study in India found that women preferred the emotional support of a family member (generally a female relative, especially mothers-in-law or mothers) while delivering in facilities (Bhattacharyya et al., 2013). While most of the past research on the impact of support for women has been in the developed world, including all of the studies in the systematic review mentioned above, there is increasing awareness in the developing world, including India, on the importance of support during childbirth. For example, the government of the state of Tamil Nadu issued an order entitling women in government hospitals the right to have one companion in the delivery room. Qualitative interviews with women and support companions in Tamil Nadu, found that they felt that this would help protect them and their loved ones from abuse in health facilities; however, people were not aware of the government order allowing family members to be present (Sri, 2009). One intervention, called Yashoda, aimed at increasing support for women in India, placed a support worker or birth companion in facilities with high delivery volumes. The Yashoda intervention has been shown to improve exclusive breast feeding, family planning uptake, nutrition, immunisation, identification of danger signs and various aspects of postnatal care such as blood pressure and temperature checks of the infants (Varghese et al., 2014). Further research is needed to identify how support during childbirth is associated with mistreatment in order to improve women's experiences. Although the growing body of literature has noted high prevalence of mistreatment during childbirth in facilities, there has been no known quantitative research that measures the association between support (who was with a woman at the time of childbirth), what type of support that person provided, and the level of mistreatment a woman experienced. Few studies have attempted to disentangle what components of support are most helpful to women, and who specifically can provide these types of support most effectively. Qualitative interviews from women who experienced an infant death in New Delhi found that women felt that prohibiting partners from being allowed in the room for antenatal care or childbirth led to difficulty understanding what the doctor was telling them about their care and to problems convincing their partners to allow them to take the actions that the doctor instructed (Saikia et al., 2015). Another qualitative study from this study setting (Lucknow) included women who recently delivered, and described how husbands and mothers-in-law were yelled at and not allowed in the room with the woman during childbirth (Sudhinaraset et al., 2015). In the Indian context, female family members, usually mothers or mothers-in-law, accompany women to health facility during antenatal care and childbirth, but
115
are not allowed in the delivery room for reasons such as to avoid crowding. Increasingly, husbands are accompanying women to the facility for childbirth as they are expected to do tasks such as arranging transport, or getting blood and medicines. Their consent is also needed to undertake caesarean if the situation arises. For the last five years, a community level health worker (Accredited Social Health Activists or ASHAs), appointed by government, is expected to accompany rural women to health facilities for childbirth, for which she gets paid. Studies have not quantitatively assessed the type of support that different family members and close ties provide, and how it influences experiences of mistreatment. Social support is typically divided into three types of support: (1) emotional support, typically provided by a close intimate tie; (2) instrumental support, referring to aid or assistance of needs such as cooking, cleaning, paying bills; and (3) informational support (Berkman et al., 2000). Qualitative findings suggest that husbands and mother/mothersin-law may play different roles during the childbirth process. For example, husbands may be more likely to provide instrumental support, such as paying for bills and buying women needed medication, whereas mother/mothers-in-law provide emotional support during and after childbirth (Sudhinaraset et al., 2015). It appears that family members could potentially be important sources for women to get information and be able to act on that information, but there are other types of support, including emotional or instrumental (bringing the woman water, supplies) that might also be important. It is important to understand what types of support women are lacking and who currently provides specific type of support to women, given that women might be supported by their husbands, mothers or mothers-in-law, other family members, local health workers, or providers in the facility in this setting. Family members and health professionals can provide support before, during, and after childbirth, including both post partum in the facility and at home. Especially in India, where women are incentivized to stay in facilities for 48 hours after childbirth, their family members could provide important care and also help identify adverse outcomes if they arise in the postpartum period. Furthermore, it is important to understand if specific supporters are more protective of women experiencing mistreatment, not only adverse clinical outcomes.
Data and methods Data collection Data was collected from a cross-sectional sample in Lucknow, Uttar Pradesh, India in May, 2015 from a total of 759 young women living in economically disadvantaged (slum) areas. Eligibility criteria included having given birth in the previous five years and being aged 16–30 years at the time of the survey. This age range was chosen as we were particularly interested in whether younger women are more at risk of experiencing mistreatment and these are the prime childbearing years. An initial sampling frame of households in the slum areas was constructed, and from that, households were approached to determine if an eligible woman lived in that household. This analysis is limited to the 392 women from that sample who had delivered in a health facility (not at home). Household surveys were administered by four trained research assistants and covered a broad scope of topics including demographic characteristics, migration experiences, fertility, pregnancy, and childbirth experiences. Respondents were interviewed in their home or in another private setting selected by the respondent. Verbal informed consent was obtained from all study
116
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
participants due to the low literacy levels in these communities. All study documents were reviewed and approved by the Institutional Review Boards at the University of California, San Francisco, and Foundation for Research in Health Systems (FRHS) in India. Variable creation Mistreatment score We constructed an additive index score of mistreatment in health facilities using a series of questions from the survey regarding specific forms of mistreatment. The included questions asked women whether or not they experienced discrimination, physical or verbal abuse, threats to withhold treatment, lack of information, abandonment, their choice of position denied, requests for bribes, or unnecessary separation from the infant. These are subjective measures of a women's perception of her treatment. Answering yes to a question resulted in a score of 10 whereas answering no to a question resulted in a score of 0. If the respondent could not answer yes or no this was converted to a missing response. The total score any woman received was divided by the number of questions that they answered. Therefore, the mistreatment score of a woman who answered all the questions was comparable to that of a woman who was only able to answer some of the questions. Each woman received a score ranging from 0 to 10. The overall mean mistreatment score was 1.87 (SD 2.86). Support score Women were asked a series of questions about whether they received certain types of support and who provided that support, both during childbirth and in the postpartum period. This analysis focuses on the support provided during childbirth. Women were asked if someone did the following: brought them water or food; helped them talk to providers; provided them information about the infant; provided them support during labour; or provided encouragement during labour. While this might not be an exhaustive list of possible types of support, we felt it a representative sample of important areas of support, based on previous qualitative research on this subject in this setting. Women were then asked to indicate which people provided each type of support, and they could list more than one. Options included: mother-in-law, mother, husband, sister, friend/neighbour, ASHA/USHA, Anganwadi worker (community based health worker), Other health worker, and “other.” Specific “other” options commonly included brother, brother-in-law, father, father-in-law, and sister, which were coded as the “friend/neighbour” category, to expand this to “friend/neighbour/other family member”. Respondents also commonly mentioned doctors and nurses, who were recoded as “other health workers.” We also constructed a support score that combined the 5 components of support into a score (0–5). The first analysis presents women who reported that no one provided each type of support, and the second and third look at whether women did have support for each component, and who provided that support. Demographic and household variables All demographic variables were self-reported. Age was grouped into three categories: 16–19; 20–24; and 25–30. Women under 20 were separated, even though this makes the age ranges inconsistent, since we hypothesised that these women might be most at risk of mistreatment and lack of support. Women and their husband's education were grouped in three categories as well: no school/illiterate; primary; and secondary or more. Other variables included caste (scheduled caste, scheduled tribe, other backwards caste and no caste, as is the standard in this setting); religion
(Hindu and Muslim); number of living children; and whether the woman was currently employed. In addition, we constructed a wealth quartile variable in order to understand differences in support experiences in the context of comparative wealth among women. This variable was based on a series of survey questions regarding whether women possessed or had access to various appliances, technology, utilities, and vehicles. Examples included possessing a refrigerator, computer, running water, or a car respectively. Having answered yes to more questions would lead to a higher score; indicating a higher level of wealth. Scores were additive and the sample was divided equally into quartiles. These were based on the overall sample of 759 survey respondents and therefore, women who delivered in facilities (n¼ 392) were not equally distributed among the four quartiles. In ascending order of wealth, quartiles were labelled lowest quartile, second lowest quartile, second highest quartile, and highest quartile. Methods Three different regression analyses were run, controlling for the variables mentioned above. The first analysis examined the association between a woman reporting not having support and her reported mistreatment score (Table 4). The second analysis explored the association between a woman reporting that a specific person provided support and her reported mistreatment score (Table 5). The final analysis explored the relationship between different types of support provided by different individuals, and her reported mistreatment score. For this final analysis, we grouped the support into three categories: (1) Instrumental (bringing water or food); (2) Informational (providing information about infant or helping to talk to providers); and (3) Emotional (support or encouragement during labour) (Table 6).
Findings Basic demographics The quantitative survey included 392 women who reported that their most recent childbirth took place in a health facility in the last five years. Exactly half of the sample reported that this birth took place in a public sector primary health centre. Other reported childbirth sites were government hospitals (20.9%), private hospitals (20.9%), and private clinics (8.2%). A little more than half of women in this sample (54.7%) reported some form of mistreatment during a facility birth. The sampled women had a mean age of 25.3 years (SD ¼ 3.1 years) at the time of the survey, with 19 women (4.9%) less than 20 years of age, 134 (34.2%) aged 20–24 years, and 239 (60.9%) aged 25–30 years. Migrant women (women who had not lived in Lucknow their whole lives) were overrepresented, making up 92.9% of the sample. Approximately two-thirds of the women were Hindu (68.6%) with the remainder being Muslim. Most women belonged to a scheduled caste (33.4%), other backwards castes (40.6%) or scheduled tribe (11.7%) (the remaining had no caste). Almost the entire sample was currently married (99.5%) and reported living in households with a mean of 4.8 members (SD ¼1.7 members). While approximately a quarter of women reported achieving a secondary education or higher (25.8%), 148 women (37.8%) reported only having a primary education and 143 women (36.5%) reported having no education at all. Seventy-seven women (19.7%) fell into the poorest wealth quartile, followed by 88 (22.5%) in the lower middle quartile, 101 (25.8%) in the upper middle quartile, and 125 (32.0%) in the highest quartile.
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 1 Women who reported they had no one to provide the following types of care/ support at childbirth. N Brought them water/food Talk to providers Provide information about the infant Provide labour support Provide encouragement
Score of lack of support 0 1 2 3 4
%
1 84 63 50 38
0.3 21.4 16.1 12.8 9.7
216 115 49 11 1
55.1 29.3 12.5 2.8 0.3
117
(59.7%), followed by mothers (33.9%), sisters/brothers/fathers (including -in-laws) (28.8%) and mothers-in-laws (21.4%) (Table 3). Health workers or friends rarely brought water or food. Husbands were the most common person to help talk to providers (43.4%), followed by health workers (20.4%), mothers (29.1%), mothers-inlaw (15.8%) and other family members (11.2%). In terms of providing information about the infant, health workers were the most common person to do this (41.6%), followed by the woman's mother (34.9), and then her mother-in-law (19.4%). Support for labour and/or breathing was most often provided by health workers (54.3%), followed by mothers (30.4%) and mothers-in-law (21.2%). Finally, mothers, husbands, and health workers all provided encouragement or guidance in labour at about the same frequency (37.8%, 37.2% and 36.7%, respectively). Mothers-in-law also frequently provided support (29.8%), followed by friends (17.65%). Multivariate regressions
Experiences of support Support during childbirth Only one woman reported that she had no one to bring her water during childbirth (Table 1). Eighty-four women (21.4%) said they had no one to help them talk to providers, 63 (16.1%) said they had no one to provide information about the infant, 50 (12.8%) had no one to provide support in labour and 38 (9.7%) had no one to provide encouragement in labour. Overall, 55.1% (216) women reported someone provided all types of support, 29.3% had no one for 1 type of support, 12.5% for two types, 2.8% for three types and 0.3% for 1 type.
Lack of support during childbirth and mistreatment Having no one to talk to providers (2.192, p o0.01) or provide information about the infant (1.606, p o0.01) was associated with a higher mistreatment score, after controlling for the demographic factors (Table 4). Reporting more experiences of lack of support overall was also associated with reporting more mistreatment (0.907, p o0.01). Being from scheduled tribe or other backwards caste was consistently associated with higher mistreatment scores, compared to women from other castes. Additionally, being from the highest wealth quintile (compared to lowest) was associated with reporting higher mistreatment scores for all types of lack of support.
Person who provided support Most women, 77.8% (305) women reported that their husband provided them support for at least one of the indicators (Table 2). Most women, 80.4% (315), said that their mother-in-law or mother provided them support for at least 1 indicator, 59.9% (235) that their sister, friend or other, and about 52.8% (207) that a health worker such as an ASHA, anganwadi or other provided this support.
Person who provided support and mistreatment Reporting that her husband ( 1.384, po 0.01) or a health worker ( 0.879, p o0.01) was the person who provided at least one type of support during childbirth was associated with a lower mistreatment score (Table 5). Reporting that her mother or mother-in-law or a sister, friend or other person provided at least 1 type of support was not associated with the mistreatment score.
Type of support provided by each person Water and/or food was most often supplied by husbands Table 2 Women who indicated that they had support from the following people for at least 1 aspect of care/support during childbirth.
Husband Health worker (ASHA, anganwadi, other) Sister, friend, other Mother-in-law, mother
N
%
305 207 235 315
77.8 52.8 59.9 80.4
Types of support provided by specific people and mistreatment No types of support provided by a woman's husband were associated with the mistreatment score (Table 6). Having either their mother or mother-in-law provide informational support was significantly associated with the woman reporting a lower mistreatment score ( 0.835, p o0.05). Reporting that a health worker provided emotional support was associated with a lower reported mistreatment score ( 0.723, p o0.05). Different types of support from other family members or friends had different associations with mistreatment. An other family member or friend providing emotional support was associated with increased mistreatment scores (1.409, p o0.01), but one of these people providing
Table 3 Percent of women reporting types of support provided by each provider of support.
MIL Mother Husband Sister/Brother/Father (including in-laws) Friend All Health Workers
Instrumental
Informational
Emotional
Brought food or Water (%)
Helped Talk to providers (%)
Info about infant Support for labour, (%) breathing (%)
Encouragement or guidance (%)
21.4 33.9 59.7 28.8
15.8 29.1 43.4 11.2
19.4 34.9 7.1 7.4
21.2 30.4 0.8 2.8
29.8 37.8 37.2 8.4
7.9 2.6
5.6 20.4
8.2 41.6
9.4 54.3
17.6 36.7
118
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 4 Association between mistreatment score and lack of support during childbirth (Odds Ratio, SD). No one to pro- No one to help vide water talk to providers
No one to Provide information about the infant
No one to provide labour support
No one to provide encouragement
Score of lack of support during childbirth
0.857 (2.799)
2.192n (0.348)
1.606n (0.404)
0.248 (0.451)
0.237 (0.505)
0.907n (0.181)
0.788 (0.692)
0.643 (0.658)
0.561 (0.680)
0.779 (0.692)
0.769 (0.694)
0.629 (0.670)
25–30 years
0.256 (0.698)
0.190 (0.662)
0.177 (0.683)
0.237 (0.698)
0.229 (0.699)
0.193 (0.675)
Number of Living Children
0.111 (0.142)
0.127 (0.135)
0.111 (0.139)
0.104 (0.143)
0.107 (0.142)
0.111 (0.138)
0.245 (0.355)
0.392 (0.338)
0.207 (0.347)
0.228 (0.355)
0.249 (0.355)
0.185 (0.343)
0.00988 (0.452)
0.0263 (0.429)
0.0748 (0.442)
0.00843 (0.452)
0.00810 (0.452)
0.0922 (0.437)
0.0472 (0.366)
0.191 (0.347)
0.110 (0.359)
0.0296 (0.365)
0.0374 (0.365)
0.00344 (0.353)
Secondary or more
0.397 (0.412)
0.576 (0.393)
0.468 (0.404)
0.391 (0.412)
0.399 (0.412)
0.469 (0.399)
Currently works outside the home
0.162
0.0715
0.161
0.171
0.172
0.0469
(0.318)
(0.304)
(0.311)
(0.317)
(0.318)
(0.308)
0.185
0.0698
0.0744
0.19
0.191
0.111
(0.338)
(0.321)
(0.331)
(0.338)
(0.338)
(0.326)
0.286 (0.487)
0.366 (0.463)
0.202 (0.477)
0.284 (0.487)
0.314 (0.490)
0.184 (0.471)
Scheduled Tribe
1.769n (0.590)
2.017n (0.561)
1.276† (0.590)
1.703n (0.602)
1.818n (0.598)
1.094§ (0.586)
Other backwards caste
1.199† (0.474)
1.258n (0.450)
0.925† (0.468)
1.186† (0.473)
1.211† (0.475)
0.978† (0.460)
Migrant (compared to nonmigrant)
0.0223
0.343
0.211
0.0213
0.0212
0.262
(0.566)
(0.541)
(0.557)
(0.566)
(0.566)
(0.550)
0.387 (0.446)
0.220 (0.423)
0.334 (0.436)
0.360 (0.446)
0.400 (0.447)
0.110 (0.434)
0.534 (0.442)
0.557 (0.420)
0.516 (0.433)
0.513 (0.444)
0.548 (0.443)
0.388 (0.429)
Mistreatment Score
Age group (compared to 16–19 years) 20–24 years
Woman's Education level (compared to no education/ illiterate) Primary
Secondary or more
Man's Education level (compared to no education/ illiterate) Primary
Religion (Muslim compared to Hindu)
Caste (compared to no caste) Scheduled Caste
Wealth quartile (lowest quartile) Second-lowest quartile
Second highest quartile
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
119
Table 4 (continued ) No one to pro- No one to help vide water talk to providers
No one to Provide information about the infant
No one to provide labour support
No one to provide encouragement
Score of lack of support during childbirth
Highest quartile
1.753n (0.462)
1.316n (0.444)
1.513n (0.456)
1.726n (0.464)
1.766n (0.462)
1.225n (0.459)
Constant
0.464 (1.113)
0.547 (1.057)
0.726 (1.092)
0.467 (1.113)
0.48 (1.114)
0.654 (1.077)
Observations R-squared
380 0.109
380 0.197
380 0.146
380 0.11
380 0.11
380 0.167
n
† §
po 0.01. p o 0.05. p o 0.1.
informational support was associated with lower reported mistreatment scores ( 0.855, po 0.05). None of the other types of support were associated with mistreatment scores for any people providing the support. Interestingly, instrumental support provided by any of the sources of support was not associated with the mistreatment score.
Discussion This is one of the first studies to explore in detail both who provides support to women and what type of support is provided by different people who accompany women during childbirth at facilities. Taking this one step further, we were also able to then see how lack of support overall and specific types of support being provided by specific sources of support (people), were associated with women's experiences of mistreatment. The majority of women in the sample had some form of support for all measures. The most common type of support that women were lacking during childbirth was having someone to help them talk to providers (21%). This is relevant in respect to past qualitative research where women discussed not being able to understand what the providers wanted them to do and how their partners being refused access made it harder to act on doctors' advice (Saikia et al., 2015). About 1 in 10 women reported not having anyone to provide labour support or encouragement. While this is a minority of the women, it still means that a substantial number of women are delivering without the emotional support that past literature has suggested is associated with improved physical and mental health outcomes. Interestingly, mothers-in-laws/mothers (80%) and husbands (79%), were the most common people to provide support, and health workers only provided support to about half (53%) of women. It appears that despite restrictions on physical accessibility to the woman during childbirth, family members are still the most frequent provider of support to women. It is also alarming that women do not report that health care providers are providing informational support to them in this period given that providers should be trained in counselling women on health care processes. It is possible that providers are simply not giving women information about what is happening to them during their labour and delivery, perhaps due to limited time, skill or willingness. However, it is also possible that providers are providing information to women in a way that women are not able to absorb or understand, perhaps due to the use of highly technical terms, not enough time for women to ask questions or providers to make sure women understand, or other barriers. A significant literature has described system-level challenges in supporting healthcare
providers to deliver the best care (Jewkes et al., 1998; Kruk et al., 2010; Freedman and Kruk, 2014). The failure of health systems for women during childbirth occurs at every level – from the facility to national and policy level (Freedman and Kruk, 2014). Providers may be mistreating women because they themselves do not have the time, resources, or system support needed to effectively perform their work duties. Jewkes et al. (1998) provides an overview of nurse–woman relationships in South African obstetric care, and finds that humiliation and physical abuse of women was quite common in this setting (Jewkes et al., 1998). Explanations for these behaviours included system level challenges such as nurses struggling to assert their power among lower class women, lack of accountability of services and lack of quality assurance by facility managers and higher-level providers, and overall culture of abuse in nursing care. Improving quality assurance and accountability at the highest levels may improve normative behaviours at the facility level to improve the care of women during labour and delivery (Freedman and Kruk, 2014). We find that lower caste women are more likely to report mistreatment (discussed in more detail in a forthcoming paper focusing on this from the same data set), suggesting that power dynamics in terms of the interaction between women and health facility staff may be at play in this setting as well (Sudhinaraset et al., 2016). Even though only about half of women reported having support from a health worker, it appears that the support of these people is important for reducing the experience of mistreatment. When health care providers are playing a positive role in women's childbirth experience, such as by providing information or encouragement, this improves women's overall view of her childbirth experience. Health care providers are already in these facilities, so engaging them more in providing types of support that women are currently lacking could have a big impact on overall experiences. However, these health care providers are often overworked and underpaid, especially in public facilities, therefore, it could prove challenging to encourage them to show more care to women at the time of childbirth without providing them additional support (World Health Organization, 2006). A landscape analysis of factors contributing to abuse and disrespect in childbirth highlighted factors at various levels, including individual/community; national laws and policies, human rights and ethics; governance and leadership; service delivery; and providers (Bowser and Hill, 2010). They discussed in detail how lack of standards, poor accountability, provider demoralisation, shortages of human resources and lack of professional development opportunities and low provider status and respect all contribute to abuse and disrespect. Lack of informational support (no one to help talk to providers or provide information about the infant) was most strongly
120
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 5 Association between mistreatment score and person providing support during childbirth (Odds Ratio, SD). Husband Health worker
Mistreatment score
Age group (compared to 16–19 years) 20–24 years
1.384 (0.344)
n
0.879 (0.300)
Sister/ Friend/Other
n
Mother/ Mother-inlaw
0.472 (0.298)
0.361 (0.375)
Table 5 (continued ) Husband Health worker
Sister/ Friend/Other
Mother/ Mother-inlaw
Second highest quartile
0.606 (0.433)
0.602 (0.438)
0.538 (0.441)
0.537 (0.442)
Highest quartile
1.652n (0.452)
1.747n (0.456)
1.802n (0.461)
1.722n (0.462)
0.732 (0.677)
0.846 (0.684)
0.841 (0.691)
0.780 (0.691)
Constant
1.932§ (1.149)
0.948 (1.113)
0.276 (1.115)
0.796 (1.167)
25–30 years
0.208 (0.683)
0.258 (0.689)
0.230 (0.695)
0.226 (0.697)
Observations R-squared
380 0.147
380 0.130
380 0.115
380 0.112
Number of Living Children
0.107
0.0930
0.108
0.111
n
§
(0.139)
Woman's Education level (compared to no education/illiterate) Primary
(0.142)
0.241 (0.354)
0.0594 0.114 (0.442) (0.448)
0.00844 (0.450)
0.0357 (0.452)
0.00742 (0.357)
0.181 (0.364)
0.00408 (0.365)
0.0320 (0.365)
Secondary or more
0.457 (0.404)
0.379 (0.408)
0.473 (0.414)
0.406 (0.412)
Currently works outside the home
0.00270
0.112
0.186
0.155
(0.313)
(0.314)
(0.317)
(0.317)
0.203
0.156
0.218
0.197
(0.330)
(0.333)
(0.337)
(0.337)
0.235 (0.477)
0.399 (0.483)
0.318 (0.486)
0.307 (0.487)
Scheduled Tribe
1.708n (0.577)
1.947n (0.586)
1.863n (0.590)
1.665n (0.599)
Other backwards caste
1.073† (0.464)
1.406n (0.474)
1.215† (0.472)
1.184† (0.473)
Migrant (compared to non-migrant)
0.102
0.167
0.0243
0.0458
(0.555)
(0.563)
(0.565)
(0.566)
0.376 (0.435)
0.307 (0.441)
0.320 (0.445)
0.354 (0.445)
Religion (Muslim compared to Hindu)
Caste (compared to no caste) Scheduled Caste
Wealth quartile (lowest quartile) Second-lowest quartile
p o 0.01. po 0.05. p o0.1.
(0.142)
0.284 (0.355)
Man's Education level (compared to no education/illiterate) Primary
0.265 (0.347)
(0.141)
0.443 (0.357)
Secondary or more
†
associated with reporting a higher mistreatment score. On the other hand, instrumental and emotional support were not associated with higher levels of mistreatment (except increased mistreatment reported when emotional support was provided by other family or friends). This suggests that women who understand what is happening during childbirth, including decisions made about their care, have better experiences during childbirth. This finding supports broader calls for person-centred care approaches, which highlights the need to include women in the decision-making process and inform women on all aspects of care. Husbands were the person most frequently reported as providing instrumental support (getting water), although other family members, including mothers and mothers-in-law also provided this type of support. The importance of husbands’ presence and support in improving women's experiences in childbirth has been shown elsewhere. For example, past qualitative research in India also found that women described their husbands as providing instrumental support, such as helping arrange transportation and buy supplies when needed for childbirth (Sudhinaraset et al., 2015). Research in Nepal found that women reported higher scores on a measure of agency during labour and delivery when their husbands were with them than when female friends were with them (although this was also associated with higher scores on the agency scale overall) (Sapkota et al., 2012). In our study, husbands were the most frequent person recorded as having helped the woman talk to providers, although mothers also provided this type of support. The provision of this type of support could explain previous findings about increased feelings of agency with husband's presence. Informational support (providing information about the infant) was much more frequently provided by health workers and mothers/ mothers-in-law and rarely by husbands. This is to be expected since husbands might be much less likely to be as deeply involved in infant-care. Women reported that health providers were common sources of support and encouragement during labour, as were mothers and mothers-in-law. Husbands provided encouragement, however, they were not frequently reported as providing support during labour. This is intuitive since the support question was about specific types of support (with breathing or techniques in labour), and husbands might have less knowledge about these very female-specific needs, although male partners in other settings have been able to provide women this type of support. Mothers and mothers-in-law provided the highest level of
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
Table 6 Association between types of support provided and mistreatment score, by person providing the support (Odds Ratio, SD).
Instrumental Support
Emotional Support
Informational Support
Husband Mother/Mo- Health ther-in-law worker
Other Family member /Friend
0.422 (0.309)
0.344 (0.305)
0.104 (0.311)
1.128 (0.901) n
Table 6 (continued ) Husband Mother/Mo- Health ther-in-law worker
Other Family member /Friend
0.443 (0.446)
0.213 (0.448)
0.298 (0.439)
0.400 (0.435)
Second highest quartile
0.569 (0.440)
0.518 (0.441)
0.664 (0.437)
0.764§ (0.433)
Highest quartile
1.654† (0.462)
1.626† (0.463)
1.632† (0.456)
1.698† (0.450)
Wealth quartile (lowest quartile) Second-lowest quartile
†
0.391 (0.312)
0.347 (0.331)
0.723 (0.352)
1.490 (0.345)
0.407 (0.322)
0.835n (0.348)
0.427 (0.345)
0.855n (0.355)
0.744 (0.690)
0.920 (0.682)
0.798 (0.674)
Constant
1.406 (1.175)
0.805 (1.154)
1.175 (1.113)
0.252 (1.086)
Observations R-squared
380 0.128
380 0.124
380 0.142
380 0.162
Age group (compared to 16–19 years) 20–24 years 0.765 (0.692)
25–30 years
0.342 (0.695)
0.207 (0.696)
0.470 (0.691)
0.109 (0.682)
Number of Living Children
0.124
0.0841
0.115
0.0696
n
0.0559 (0.445)
0.0384 (0.441)
0.0121 (0.363)
0.0375 (0.365)
0.143 (0.362)
0.121 (0.357)
Secondary or more
0.437 (0.410)
0.434 (0.411)
0.444 (0.406)
0.537 (0.405)
Currently works outside the home
0.0598
0.179
0.0739
0.183
(0.317)
(0.316)
(0.314)
(0.309)
0.185
0.203
0.119
0.0495
(0.336)
(0.336)
(0.332)
(0.330)
0.149 (0.492)
0.302 (0.485)
0.310 (0.480)
0.332 (0.478)
Scheduled Tribe
1.700† (0.601)
1.547n (0.602)
1.873† (0.583)
1.782† (0.575)
Other backwards caste
1.089n (0.473)
1.137n (0.473)
1.172n (0.468)
0.964n (0.467)
Migrant (compared to non-migrant)
0.191
0.0849
0.196
0.121
(0.574)
(0.564)
(0.561)
(0.555)
Caste (compared to no caste) Scheduled Caste
p o 0.05. po 0.01. p o 0.1.
(0.139)
0.0431 0.0343 (0.449) (0.450)
Religion (Muslim compared to Hindu)
0.250 (0.354)
(0.141)
0.262 (0.348)
Woman's Education level (compared to no education/illiterate) Primary
0.248 (0.353)
(0.142)
0.368 (0.352)
Secondary or more
† §
(0.141)
Woman's Education level (compared to no education/illiterate) Primary
121
support across the different categories (instrumental, informational and emotional). A number of studies in South Asia have highlighted the gate-keeper role of mothers-in-laws in reproductive health care decision-making and practices, including antenatal care use, place of childbirth, family planning use, abortions, and postnatal practices including breast feeding (Barua and Kurz, 2001; Ganatra and Hirve, 2002; Masvie, 2006; Char et al., 2010; Diamond-Smith et al., 2012; Sudhinaraset et al., 2015). However, few studies have documented the support that mothers and mothers-in-law are providing to women during childbirth. One study on Kangaroo Mother Care (KMC) in Chandigarh, India found that mothers-in-law provided KMC for over half of women (Parmar et al., 2009). It is clear that a better understanding is necessary of what role mothers and mothers-in-law play at the facility in providing care and support to women, especially for planning interventions that might want to target these already existing forms of support to engage them in providing additional support for women. Women who received support from husbands and health workers had lower mistreatment scores, while there was no association between mistreatment and receiving support from mothers/mothers-in-law and other family members/friends. Given the patriarchal nature of Indian society, it seems plausible that health providers treat women with more respect when their husbands are actively engaged in their childbirth and that the wishes and desires of women are more often fulfilled if their husband is supporting them and advocating for them. This suggests that altering policies to encourage (or allow) husbands/ partners to be able to accompany women could reduce experiences of mistreatment. It is possible that providers, who are often female (in Indian most providers for women during childbirth at all levels are female), might also be more afraid of abusing, disrespecting, or demanding bribes of women (components of the mistreatment score) when their husbands are present. Although husbands frequently provided instrumental support, this was not associated with the woman reporting a lower mistreatment score. Mothers or mothers-in-law providing informational support (helping communicate with providers or providing information about infant care) was significantly associated with
122
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
lower mistreatment scores, and these individuals were frequently reported as providing these types of support although they were not the most common person for either category. The only person whose provision of emotional support was significantly associated with lower mistreatment scores was a health worker. One study (in the US) looking at both student midwives and laywomen as support for women during childbirth found that women were satisfied with both types of people playing supportive roles (Hemminki et al., 1990). Women who did not have their husband with them were especially positive about having health workers as support. However, the student midwives did not see much value in playing this role for women and were generally negative about the trial. In our study, contrary to the situation with health providers, having emotional support from other family members or friends was associated with higher mistreatment scores. This is in contrast to past literature, which found that support from a friend or close family member was associated with more satisfaction, which, although a different measure, is associated with mistreatment (Hodnett et al., 2012). It is possible that emotional support is needed from other family or friends when the childbirth experience is especially bad. In other words, it is possible that the relationship goes the other way, where the experience of mistreatment leads women to call upon their family or friends to provide them emotional support. There are several limitations to this study. First, data was only collected from women ages 16-30 living in slum areas of Lucknow, and therefore is unlikely to be generalisable to richer women in Lucknow or to women in other parts of Uttar Pradesh or India. Furthermore, migrant women were overrepresented in the survey, therefore, these findings might not be reflective of non-migrant women. All data was collected retrospectively, as women had delivered in the last 5 years, and therefore might be subject to recall bias. Our measure of mistreatment is based on women's selfreports of their own experiences of mistreatment, which is a subjective measure and thus the same experience might be interpreted differently by different types of women. We attempt to control for this by including variables in the analysis such as education and wealth, as past literature has suggested that richer and more educated women report higher levels of mistreatment. We are aware that women may interpret the meaning of the wording about “information” or “respect” differently, however, our interest is a woman's experience (perception, internalisation), rather than what actually happened. Future research could use mixed methods combining these measures with observations, in order to triangulate findings on perceived and actual experiences. Also related to women's perceptions of mistreatment, it is possible that women that had more adverse post partum outcomes might have had more unfavourable opinions of their care looking back. We were unable to analyse any associations between mistreatment and adverse maternal or child health outcomes, due to the small sample size and relative rarity of these events. We recommend that future research explore how mistreatment is associated with outcomes such as complications in labour and use of postnatal services in this population. We also suggest that research in westernized countries focused more on women's experiences of childbirth, in addition to clinical and behavioural outcomes. The mounting body of evidence on mistreatment during childbirth, which this research adds to, needs to be funneled into interventions aimed to reduce mistreatment and improve women's childbirth experiences. Support for women is especially important because this is a component of mistreatment that is associated with improved outcomes and behaviours. Very little attention has been paid to how support and mistreatment are interlinked in global contexts, and especially in low and middleincome countries, despite the attention on mistreatment broadly. This research provides an important contribution in that it not
only looks at the relationship between the absence of specific types of support (instrumental, informational and emotional) and mistreatment, but it takes this one step further by looking at what specific people provide support, and when the support of specific people for specific domains actually are most able to deter mistreatment. We recommend that interventions be developed that turn these findings into action by engaging different family members and health workers to provide specific types of support to women during childbirth. Utilising family members and health workers to provide support provides a low-hanging fruit, since these individuals are already often with women at childbirth, and thus need only be incentivized and engaged to uphold the professional standards of practice, quality monitoring and regulation. We also recommend that institutional policy regarding support people for women during childbirth be expanded to acknowledge the importance of and increase support to women during childbirth, and highlight the necessity of allowing multiple people to provide different forms of support to women.
Acknowledgements The authors would like to acknowledge the Bill and Melinda Gates Foundation (Grant ID: OPP1084330) for the funding for this project. We would also like to acknowledge Kanksha Singh, Emily Treleaven, and Rajni Wadwa.
References Barua, A., Kurz, K., 2001. Reproductive health-seeking by married adolescent girls in Maharashtra, India. Reproductive Health Matters 9, 53–62. http://dx.doi.org/ 10.1016/S0968-8080(01)90008-4. Berghella, V., Baxter, J.K., Chauhan, S.P., 2008. Evidence-based labor and delivery management. American Journal of Obstetrics and Gynecology 199, 445–454. http://dx.doi.org/10.1016/j.ajog.2008.06.093. Berkman, L.F., Glass, T., Brissette, I., Seeman, T.E., 2000. From social integration to health: Durkheim in the new millennium. Social Science & Medicine 51, 843–857. http://dx.doi.org/10.1016/S0277-9536(00)00065-4. Bhattacharyya, S., Srivastava, A., Iqbal Avan, B., 2013. Delivery should happen soon and my pain will be reduced: understanding women's perception of good delivery care in India. Global Health Action 6. http://dx.doi.org/10.3402/gha. v6i0.22635. Bohren, M.A., Vogel, J.P., Hunter, E.C., Lutsiv, O., Makh, S.K., Souza, J.P., et al., 2015. The mistreatment of women during childbirth in health facilities globally: a mixed-methods systematic review. PLoS Medicine 12 (6), e1001847, 1-32, http://dx.doi.org/10.1371/journal.pmed.1001847. Bowser, D., Hill, K., 2010. Exploring Evidence for Disrespect and Abuse in FacilityBased Childbirth: Report of a Landscape Analysis. USAID. Char, A., Saavala, M., Kulmala, T., 2010. Influence of mothers-in-law on young couples’ family planning decisions in rural India. Reproductive Health Matters 18, 154–162. http://dx.doi.org/10.1016/S0968-8080(10)35497-8. Diamond-Smith, N., Campbell, M., Madan, S., 2012. Misinformation and fear of sideeffects of family planning. Culture Health & Sexuality 14, 421–433. http://dx. doi.org/10.1080/13691058.2012.664659. Freedman, L.P., Kruk, M.E., 2014. Disrespect and abuse of women in childbirth: challenging the global quality and accountability agendas. Lancet 384, e42–e44. http://dx.doi.org/10.1016/S0140-6736(14)60859-X. Ganatra, B., Hirve, S., 2002. Induced abortions among adolescent women in rural Maharashtra, India. Reproductive Health Matters 10, 76–85. http://dx.doi.org/ 10.1016/S0968-8080(02)00016-2. Gibbins, J., Thomson, A.M., 2001. Women's expectations and experiences of childbirth. Midwifery 17, 302–313. http://dx.doi.org/10.1054/midw.2001.0263. Gruber, K.J., Cupito, S.H., Dobson, C.F., 2013. Impact of Doulas on healthy birth outcomes. The Journal of Perinatal Education 22, 49–56. Hemminki, E., Virta, A.-L., Koponen, P., Malim, M., Kojo-Austin, H., Tuimala, R., 1990. A trial on continuous human support during labor: feasibility, interventions and mothers' satisfaction. Journal of Psychosomatic Obstetrics & Gynecology 11, 239–250. http://dx.doi.org/10.3109/01674829009084420. Hodnett, E., Gates, S., Hofmeyr, G., Sakala, C., 2012. Continuous support for women during childbirth. Cochrane Database of Systematic Reviews, 10. Hulton, L.A., Matthews, Z., Stones, R.W., 2007. Applying a framework for assessing the quality of maternal health services in urban India. Social Science & Medicine 64, 2083–2095. http://dx.doi.org/10.1016/j.socscimed.2007.01.019. Jewkes, R., Abrahams, N., Mvo, Z., 1998. Why do nurses abuse patients? Reflections from South African obstetric services. Social Science & Medicine 1982,
N. Diamond-Smith et al. / Midwifery 40 (2016) 114–123
1781–1795. Kruk, M.E., Rockers, P.C., Mbaruku, G., Paczkowski, M.M., Galea, S., 2010. Community and health system factors associated with facility delivery in rural Tanzania: a multilevel analysis. Health Policy 97, 209–216. http://dx.doi.org/10.1016/j. healthpol.2010.05.002. Masvie, H., 2006. The role of Tamang mothers-in-law in promoting breast feeding in Makwanpur District, Nepal. Midwifery 22, 23–31. http://dx.doi.org/10.1016/j. midw.2005.02.003. Parmar, V.R., Kumar, A., Kaur, R., et al., 2009. Experience with Kangaroo mother care in a neonatal intensive care unit (NICU) in Chandigarh, India. Indian Journal of Pediatrics 76, 25–28. http://dx.doi.org/10.1007/s12098-009-0024-2. Saikia, N., Sehgal, S., Kumar Bora, J., Diamond-Smith, N., 2015. Why are Neonatal Deaths Higher among Urban Poor? Qualitative Insights from Metropolitan Delhi. Rev. Sapkota, S., Kobayashi, T., Kakehashi, M., Baral, G., Yoshida, I., 2012. In the Nepalese context, can a husband's attendance during childbirth help his wife feel more in control of labour? BMC Pregnancy and Childbirth 12, 49. http://dx.doi.org/ 10.1186/1471-2393-12-49. Sri, S.B., 2009. Translating medical evidence into practice: working with
123
communities and providers to promote active management of the third stage of labour. Health Popul. Innov. Fellowsh. Programme Working paper 11. Sudhinaraset, M., Beyler, N., Barge, S., Nadia, 2015. Decision-making for delivery location and quality of care among Slum-Dwellers: a qualitative study in Uttar Pradesh, India. Rev. BMC Pregnancy Childbirth. Sudhinaraset, M., Treleavan, E., Melo, J., Singh, K., Diamond-Smith, N., n.d. Women's Status and Experiences of Mistreatment during Childbirth in Uttar Pradesh: A Mixed Methods Study using Cultural Health Capital Theory. Rev., 2016. Varghese, B., Roy, R., Saha, S., Roalkvam, S., 2014. Fostering maternal and newborn care in India the Yashoda Way: does this improve maternal and newborn care practices during institutional delivery? PLoS One 9. http://dx.doi.org/10.1371/ journal.pone.0084145. Vogel, J., Bohren, M., Tunçalp, Ò, Oladapo, O., Gülmezoglu, A., 2015. Promoting respect and preventing mistreatment during childbirth. BJOG: An International Journal of Obstetrics and Gynaecology . http://dx.doi.org/10.1111/ 1471-0528.13750, n/a–n/a. World Health Organization, 2006. Working Together for Health: World Health Report 2006. The World Health Organization, Geneva.