Introduction
Every year, more than a quarter of a million women die from pregnancy-related causes, and nearly 3 million babies die before they reach 1 month of age (Lawn et al., Reference Lawn, Blencowe, Oza, You, Lee and Waiswa2014; WHO, 2015). In addition, for every mother and baby who die, there are many more who survive life-threatening complications during and after pregnancy and delivery (Say et al ., Reference Say, Souza and Pattinson2009). A disproportionate number of these deaths and ‘near-misses’ occur in low-resource settings (WHO, 2014), including sub-Saharan Africa.
In Ghana, West Africa, the maternal mortality ratio has been estimated at 319 per 100,000 live births, placing a woman’s lifetime risk of maternal death at 1 in 74, compared with 1 in 3800 in the United States (WHO, 2016). Sixty-eight per cent of all deaths among children under the age of five in Ghana happen before a child’s first birthday, with 48% occurring in the first month of life (Ghana Statistical Service et al., 2015).
While the clinical causes of such events are increasingly well-understood – including obstructed labour, hypertensive disorders, haemorrhage and complications of unsafe abortion for women (Koblinsky et al., Reference Koblinsky, Moyer, Calvert, Campbell, Campbell and Feigl2016) and birth asphyxia, sepsis, congenital malformations and complications from prematurity among newborns (Lui et al., Reference Lui, Oza, Hogan, Chu, Perin and Zhu2016) – what is less clear is how women themselves understand the aetiology of illness for mothers and babies. Previous research has described spiritual beliefs as playing a prominent role in illness attribution, especially the concept of the evil eye as a source of illness for pregnant women and their newborns (Dako-Gyeke et al., Reference Dako-Gyeke, Aikins, Aryeetey, Mccough and Adongo2013; Alexander et al., Reference Alexander, Mustafa, Emil, Amekah, Engmann, Adanu and Moyer2013; Nyambura et al., Reference Nyambura, Nyamache and Gechiko2013).
Researchers in Kenya (Matsuyama et al., Reference Matsuyama, Karama, Tanaka and Kaneko2013; Wanjohi et al., Reference Wanjohi, Griffiths, Wekesah, Muriuki, Muhia and Musoke2017), Nigeria (Adanikin et al., Reference Adanikin, Onwudiegwu and Akintayo2014; Sina et al., Reference Sina, Iyabo and Ayodele2014), Ethiopia (Amare et al., Reference Amare, Degefie and Mulligan2012; Gebrehiwot et al., Reference Gebrehiwot, San Sebastian, Edin and Goicolea2014), Gambia (O’Neill et al., Reference O’Neill, Clark and Grietens2017) and Ghana (O’Kyere et al., Reference O’Kyere, Tawiah-Agyemang, Manu, Deganus, Kirkwood and Hill2010; Adusi-Poku et al., Reference Adusi-Poku, Edusei, Bonney, Tagbor, Nakua and Otupiri2012; Dako-Gyeke et al., Reference Dako-Gyeke, Aikins, Aryeetey, Mccough and Adongo2013; Otoo et al., 2014) have all alluded to the prominence of the evil eye amidst inquiries into other aspects of maternal and newborn health and health care seeking, but none of those studies has focused exclusively on the evil eye. Thus very little is known about how the evil eye is perceived, whether perceptions differ across regions, and what is perceived to prevent or treat the evil eye.
This study sought to explore the evil eye across three separate regions of Ghana – Greater Accra, Central and Ashanti regions. It had the following aims: 1) to determine whether – unprompted – the evil eye was mentioned as a potential cause of illness by women whose newborns had survived a life-threatening complication; 2) to determine commonalities of the evil eye described across three regions of Ghana; and 3) to determine prevention and treatment strategies for the evil eye.
Methods
Study location
This was a cross-sectional, qualitative study utilizing a semi-structured interview guide. The research was conducted as part of a larger study exploring maternal and neonatal near-misses at three tertiary care hospitals in southern Ghana between March and August 2015 (Oppong et al., Reference Oppong, Bakari, Bell, Bockarie, Adu and Turpin2019). This study and its larger parent study were conducted at the Korle-bu Teaching Hospital in Accra, Komfo Anokye Teaching in Kumasi and Cape Coast Teaching Hospital in Cape Coast.
The Korle-bu Teaching Hospital (KBTH) is located in Accra, the capital city of Ghana, with a population of about 3.2 million (Ghana Statistical Service, 2010). It is the largest referral hospital in Ghana and also serves as the teaching hospital for the University of Ghana Medical School. Annually, about 13,000 women are managed at the hospital with pregnancy-related issues, with about 11,000 deliveries and a Caesarean delivery rate of about 35%.
The Komfo Anokye Teaching Hospital (KATH) is located in Kumasi, the second largest city in Ghana, with a population of about 2 million (Ghana Statistical Service, 2010). It is the referral centre for most of the mid-portion of Ghana. It is also the teaching hospital affiliated with the Kwame Nkrumah University of Science and Technology School of Medical Sciences. The annual delivery in KATH is about 11,000, with a Caesarean section rate of 32%.
The Cape Coast Teaching Hospital (CCTH) is located to the west of Accra in the coastal town of Cape Coast, about 100 miles from Accra. It serves as the main referral hospital for most of the rural Central and parts of the Western regions of Ghana. The hospital also serves as the teaching hospital of the University of Cape Coast, School of Medical Sciences (UCC-SMS). The hospital has 369 beds and conducts about 3600 deliveries per year.
Study participants
As part of a larger study (Oppong et al., Reference Oppong, Bakari, Bell, Bockarie, Adu and Turpin2019), all pregnant and newly delivered women who had experienced a maternal near-miss (as defined by the World Health Organization’s (WHO) Maternal Near Miss Screening Tool; Souza et al., Reference Souza, Cecatti, Haddad, Parpinelli, Costa and Katz2012) and newborns admitted to the hospital who experienced a neonatal near-miss (as defined by an adapted Neonatal Near-Miss Screening tool) during the study period were recruited to participate in the larger study, alongside matched healthy controls. Women and newborns were identified for this larger study using chart reviews, interviews with providers and interviews with women themselves. For this smaller qualitative portion of the study, focusing on the qualitative experiences of women whose newborns experienced a near-miss and relying upon lengthy in-depth interviews, respondents were purposively selected from the larger study to represent women of various age groups. Interviews were conducted with 20 mothers of neonates who experienced a near-miss. This included two mothers who had experienced both a near-miss themselves and whose newborns had experienced a near-miss.
Data collection
Semi-structured interviews were conducted from March to August 2015 by trained research assistants (including ZA) who approached women on the labour and delivery wards. All research assistants (three female, two male) had been through explicit qualitative methodology training, including conducting mock interviews and getting immediate feedback for improvement. After obtaining informed consent, research assistants asked open-ended questions regarding women’s perception of what happened to themselves or their newborn, what they believed caused it, how the illness was treated, and any additional factors related to their health care experience they wanted to share. The evil eye per se was not asked about directly unless the women brought it up themselves.
Interviews, which typically lasted between 30 and 60 minutes, were conducted in a staff room for privacy in Ga or Akan, then translated into English and transcribed.
Data analysis
Transcriptions were evaluated and coded by researchers using the constant comparative method of theme generation. All qualitative interviews were read by at least two of the authors (AB, ZA, JY, CM) and ‘in vivo’ coding was conducted to assist in the identification of main codes. All transcripts were entered into NVivo 10.0, and focused coding (using the initial coding structure as a guide) was conducted by four separate coders (AB, JY, ZA, CM). Coders held regular meetings to review and revise the codebook to reflect inclusion and exclusion criteria that may not have arisen previously.
The data analysis process involved repeated, lengthy discussions amongst the authors, using the identified codes to generate a framework to guide analysis. Preliminary analyses focused on describing what the evil eye means for women, what it is perceived to do to mothers and their babies, how it can be prevented and how it can be treated. Further thematic analysis focused on the bigger picture of the evil eye – attempting to determine whether there is a broader purpose the evil eye may serve within the communities. The COREQ criteria were used to guide both analysis and reporting (Tong et al., Reference Tong, Sainsbury and Craig2007).
Results
A total of 37 unique women were interviewed across the three sites, reflecting a mean age of 29.3 years (95% CI, 26.9, 31.6, range 18–42). Half of the sample was 27 or younger, another 25% was between 28 and 34 and the remaining participants were 35 or older. Women were evenly split across the three study sites. Six of the 20 mothers interviewed about their newborn’s illnesses spoke at length about the evil eye being a potential cause of the illness. Women from each site – Kumasi, Accra and Cape Coast – talked about the evil eye as a source of newborn illness without being prompted, but some spoke in greater length and with greater eloquence than others who simply mentioned it but did not elaborate. None of the seventeen women interviewed about maternal near-misses mentioned the evil eye.
What is the evil eye?
The evil eye was described in a variety of terms, including using the term ‘evil eye’ itself or mentioning evil spirits, witchcraft or describing a look from someone that results in a bad outcome for the baby. Commonalities across sites and across respondents included a person looking at a pregnant woman, her newborn baby, the baby’s clothes, and even the mother’s food, and causing harm, even unintentionally.
Hmmmm, my sister, this world is strange…some people are evil, they have bad eyes and they can make the babies sick once they visit. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
There are some who bath their babies with herbal medicine because there is a belief that if you don’t do so people with bad eyes [witches] can give babies sickness by just a glance at them. Some also do not use herbal medicine but rather keep their babies indoors so that nobody sees them. (34-year-old mother from Kumasi; ID# KA2649.34.NNM)
Several respondents described babies falling ill shortly after someone had visited, suggesting the evil eye was to blame for the baby’s condition:
Even this morning, a lady who met me at NICU had an experience like this. Her husband’s niece came to visit them at the hospital; the girl just held the hand of the baby. After she left, there was this yellowish thing on the baby’s eye. That was why her baby ended up at NICU. I told her that it is the girl who … caused this. … When I went back, I saw that the lady was crying, she told me that the baby just died. I was even there when her brother came and they wrapped the baby in polythene bag. By now they have finished burying it. …Yes… some people are evil. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
In addition, women reported that the evil eye can come from people who are not necessarily aware they have it:
It is only God who can know which one is the exact cause. (32-year-old mother from Cape Coast; ID# CC1698.32.NNM)
What can the evil eye do?
When asked what the evil eye can do to a baby, responses ranged from affecting the baby’s skin, discolouring their eyes, affecting their ability to feed, causing epilepsy or even causing death.
The sickness is called abenziwa…that is what causes the wrinkle of the babies…. the baby will not grow to be a normal human being. … If not detected early, the baby might even die. (18-year-old mother from Cape Coast; ID# CC1642.18.NNM)
The baby will be very small, as in it will grow very lean. (32-year-old mother from Cape Coast; ID# CC1698.32.NNM)
The babies [tend] to have something on their eye that turns the eye to a yellowish colour…. Also the babies [tend] not to feed very well from the breast, but when you [use] the feeding bottle, they take it. … Another characteristic is that, there will be something like a vein on their forehead…yes. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
What can be done for a baby given the evil eye?
Treating a baby affected by the evil eye typically warrants traditional medicine, as the cause is seen as spiritual, not medical.
Oooh some mothers sometimes assume that as for this sickness is not hospital sickness. They will say, ‘Ooh when I gave birth to my baby he/she wasn’t like this and so it’s either somebody has looked at the baby with an evil eye so it will be better to take the baby to someone who can see things [herbalist or traditional healer],’ so that the person can heal their baby for them because they think it is something spiritual. (Mother from Accra, age unknown; ID# KB3760.NNM)
Some respondents suggested that babies treated with allopathic medication would still benefit from seeing traditional healers, including spiritual healers.
The pastors will pray for them, to protect them from the hands of the enemy, [but] they will need the medicine from the doctors first. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
Can the evil eye be prevented?
Since people may be able to cast the evil eye on a woman or her baby without realizing it, prevention during pregnancy must be focused on the mothers covering their skin.
You should always cover your chest and neck and even you should cover the stomach itself, not only the neck. (18-year-old mother from Cape Coast; ID# CC1642.18.NNM)
All I can say is that, if you are pregnant, please cover your body and wear decent clothes, also we should be cautious of how we cater for ourselves because evil spirit is real. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
None of the respondents could explain why the chest, neck and stomach were the critical areas to cover, but these were mentioned repeatedly by several respondents. Respondents also described a common belief that pregnant women should not eat outside or in front of strangers.
There is a belief that pregnant women are not supposed to eat outside, if they do so any bad person can give sickness to their babies through the food spiritually. (31-year-old mother from Kumasi; ID# KA2609.31.NNM)
Respondents also emphasized the role of traditional medicine in not just treatment, but also prevention.
Everyone has his or her belief. Sometimes the aged say that if you are pregnant you might have some [people] who can cause harm to your baby, so if you go for such [traditional] medications or herbals it protect(s) the mother during and after pregnancy…. (34-year-old mother from Kumasi; ID# KA2649.34.NNM)
Prevention among newborns focused on keeping babies away from other people until they are old enough to survive the evil eye.
…the babies have to be indoor(s) for like a month before they are brought outside. (32-year-old mother from Cape Coast; ID# CC1531.32.NNM)
They shouldn’t give the babies to people anyhow. When they come visiting, they should say the baby is asleep. (18-year-old mother from Cape Coast; ID# CC1642.18.NNM)
Another respondent described the need to carefully wash clothes that have been hanging outside to dry, since someone passing by can impart the evil eye through the clothes that subsequently touch the baby.
When you put their clothes in blue dye [local soap], the evil spirits do not like that blue dye… Some people, after just looking, then they affect the baby. But if you are home and you see that someone just [walked by] and you don’t know the person, you just have to remove the clothes from the line and wash them again. (32-year-old mother from Cape Coast; ID# CC1698.32.NNM)
Ok so when you re-wash it, then it means the baby will not be affected by the evil spirit? (Interviewer)
Yes, and you have to put it in a blue dye and keep it aside; do not use it that day. (32-year-old mother from Cape Coast; ID# CC1698.32.NNM)
Does fear of the evil eye serve a societal purpose?
In addition to the mechanics of symptoms, prevention and treatment for the evil eye, women described the evil eye as something they could ward off through proper behaviour. Women described the evil eye as likely, ‘especially when you don’t wear any cloth to cover your chest and neck’ (18-year-old woman from Cape Coast, ID# CC1642.18.NNM). Others warned against ‘copying bad practices from friends’ and ‘being very careful in our daily activities’ (32-year-old woman from Cape Coast, ID# CC1531.32.NNM) to avoid falling victim to the evil eye. Women also advised other women to ‘be careful in their doings [and] eating habits’ (ID# CC1642.18.NNM) and not to eat outside where others can infect them by looking at their food. Women are also advised to keep themselves and their babies inside for several days to several weeks after delivery. Taken together, preventing the evil eye requires staying indoors as much as possible, dressing conservatively and behaving in what is perceived as an appropriate manner.
Discussion
Nearly a third of mothers interviewed about their baby’s illness spontaneously described the evil eye as one potential cause. The study findings indicate that the evil eye is perceived to affect women during pregnancy and delivery or newborns in the first few weeks after birth, causing a variety of maladies, including wrinkly skin, jaundice and even death among babies. Perpetrators of the evil eye may not even be aware they are casting an evil eye, and the issue of intentionality remains unclear. While some respondents reported that ‘people are evil’, suggesting that they intend to cause others harm, other respondents indicated that the evil eye can be cast unknowingly. Respondents described a variety of prevention measures, including covering the skin, keeping the baby away from others and seeking herbal or spiritual protection from traditional healers.
While the research literature addressing the evil eye is limited, these findings align with the results of other studies throughout Ghana. Dako-Gyeke et al. Reference Dako-Gyeke, Aikins, Aryeetey, Mccough and Adongo(2013) described how the evil eye is perceived to cause asram – the physical manifestation of evil spiritual attacks on a mother or her baby. Asram can affect mothers or their newborns, and in some settings it refers to any child born with cognitive or physical defects (Adusi-Poku et al., Reference Adusi-Poku, Edusei, Bonney, Tagbor, Nakua and Otupiri2012). Preventing such spiritual attacks requires women to avoid disclosing her pregnancy in the early months, not eating certain foods, not going out at night and not having a quarrelsome attitude (Dako-Gyeke, Reference Dako-Gyeke, Aikins, Aryeetey, Mccough and Adongo2013). Otoo et al. Reference Otoo, Habib and Ankomah(2015) reported that respondents described early attendance at antenatal care, such as before the pregnancy was visible, puts women at risk of exposure to evil spirits (Otoo et al., Reference Otoo, Habib and Ankomah2015). According to local beliefs in Ghana, ‘A spiritual force can destroy the fetus if it is noticed at its formation stage and this is what is locally referred to as “evil eye”’ (SEND-Ghana, 2017). As a result, women delay seeking antenatal care until they believe the fetus can survive exposure to the evil eye. Women may also forego having their mothers or sisters or other relatives or friends assist them during delivery for fear an ‘evil spirited companion’ may unintentionally cause harm.
As these results support, from a sociological standpoint, the evil eye has been described as having social control features (Nyambura et al., Reference Nyambura, Nyamache and Gechiko2013). Women are urged to cover themselves, stay home in the evenings and maintain an agreeable attitude, otherwise putting themselves at risk of evil spirits. At the same time, community members must be careful to respect social customs lest they be accused of casting an evil eye. According to Nyambura et al., to avoid being suspected of having the evil eye, when looking at a child it is necessary to say, ‘God bless it’ (Nyambura et al., Reference Nyambura, Nyamache and Gechiko2013). One noteworthy observation in the research is that the perpetrators of the evil eye were typically described as women. While none of the respondents explicitly stated that only women can cast an evil eye, more often than not descriptions were of females seeing a newborn or a pregnant woman and causing harm. This raises questions about jealousy and the premium placed on fertility and childbirth among Ghanaian women. This was alluded to by Matsuyama et al., who described dzongo as a type of evil eye caused by envy (Matsuyama et al., Reference Matsuyama, Karama, Tanaka and Kaneko2013) and Nyambura et al., who described the evil eye as a situation in which ‘envious people can cause harm by a mere glance at coveted objects or their owners’ (Nyambura et al., Reference Nyambura, Nyamache and Gechiko2013). In this case, the coveted object is a fetus or newborn baby.
While these findings corroborate the existing literature, they also complement previous findings by focusing on mothers of newborns who have survived a life-threatening complication. Mothers interviewed in a hospital setting – and thus presumably inclined to use allopathic medicine – still described the evil eye in great detail as a potential cause of their baby’s illness. These descriptions were unprompted, meaning that women mentioned the evil eye spontaneously and repeatedly across all three study sites. This finding – that the evil eye is cited as a cause of newborn illness – suggests further research is warranted to understand how the evil eye is perceived.
In addition, this study included respondents from three different regions of Ghana, representing different levels of urbanization and different ethnic composition. While exploratory qualitative studies cannot be seen as generalizable to a broader population, the inclusion of a diverse group of respondents suggests that the ‘evil eye’ is not an isolated belief in Ghana.
Despite the strengths of this study, there are weaknesses worth noting. First, since the evil eye was not the primary focus of the interviews, the interview guide did not include detailed prompts. This meant that some women mentioned the evil eye but were not further prompted, precluding equally detailed quotes from across all three study sites. It is possible that with more directed questioning alternative interpretations could have been identified or differences in how the evil eye is perceived or may manifest itself uncovered. However, the findings were consistent across respondents and dovetailed well with the existing, albeit limited, literature on this topic. In addition, the hospital-based nature of this study precluded interviews with mothers who may have only sought traditional, non-allopathic care for newborns they believed to be afflicted with the evil eye. Future research that explicitly seeks such perspectives is warranted. It is also possible that results may have been different had a site from northern Ghana been included, but this was beyond the scope of the current study.
The implications of this research are myriad. Clinicians, researchers and programmatic staff need to be mindful of perceptions of causation of illness and attribution. If it is believed that something originates with the evil eye, the type of treatment sought will not necessarily align with allopathic recommendations for care. However, it is not enough to simply tell women to come to the hospital if they observe given symptoms. Traditional beliefs are deeply rooted, and often held widely within a community. Thus community-level outreach and respectful navigation of traditional beliefs and practices is warranted.
In conclusion, the evil eye is believed to be a common source of newborn illnesses and deaths across three regions in Ghana, even among women being treated for maternal or newborn complications in allopathic health care settings. Understanding the pervasiveness of such beliefs is important for clinicians, researchers and programmatic staff in order to adequately address barriers to care-seeking and compliance with allopathic medical recommendations.
Author ORCIDs
Cheryl A. Moyer https://orcid.org/0000-0002-5796-157X
Acknowledgments
The authors would like to acknowledge the leadership at Korle Bu Teaching Hospital, Komfo Anokye Teaching Hospital and Cape Coast Teaching Hospital in Ghana for making this research possible.
Funding
Funding for this study was provided by the National Institutes of Health, via 1-D43-TW-009353-01. The NIH did not participate in the conduct of this research and is not responsible for its content.
Conflicts of Interest
The authors have no conflicts of interest to declare.
Ethical Approval
Ethical clearance was obtained from the institutional review boards at the Kwame Nkrumah University of Science and Technology for KATH (CHRPE/AP/021/15), University of Cape Coast for CCTH (UCCIRB/EXT/2015/02), University of Ghana for KBTH (MS-Et/M.7-P4.5/2014-2015) and the University of Michigan (HUM00097103). Each participant was taken through a written informed consent process, including explicit permission to audio-record the interviews.
