Background
Neonatal danger signs were proposed by World Health Organization (WHO) and United Nations International Children’s Emergency Fund(UNICEF) [
1], which indicates newborns being at high risk of illness and death. Any of these signs’ existence needs early detection. Timely and adequate care-seeking is crucial to improve neonatal health and survival. Failure to seek medical care might be related to caregivers’ inadequate knowledge of neonatal danger signs [
2]. Good knowledge of these signs plays a vital role in reducing mortality [
3]. The levels of knowledge of danger signs vary between different countries and regions. In Uganda, India, Ethiopia and Nigeria, the percentage of caregivers with poor knowledge of neonatal danger signs range from 38 to 85% [
4‐
7].
The neonatal period is one of the critical periods for a child’s survival. Every year over 2.4 million babies die during the neonatal period globally [
8]. Although the average global rate of neonatal mortality was 17 deaths per 1000 live births in 2019, down by 54% from 37 deaths per 1000 in 1990, the neonatal mortality declined slower than mortality of children aged between 1–11 months and 1–4 years [
9,
10]. The vast majority of newborn deaths occur in low and middle-income developing countries [
10].
China, with 64 thousand neonatal deaths occurred by the 2019 report, is still one of the top 10 countries with the highest number of newborn deaths in 2019 and ranks 15 in Asia in terms of neonatal mortality [
8,
11]. The neonatal mortality rate declined from 24.9 deaths per 1000 live births in 1990 to 4.5 deaths per 1000 live births in 2017 [
12]. China has made a significant progress in reducing the under-five mortality rates (Millennium Development Goal 4) by 2/3 from 1990 levels [
13]. Yet, in economically less developed regions, the risk of neonatal death is higher compared with developed ones. In rural areas, newborns have the least chance of receiving medical care before death among the population [
14,
15]. Among deaths of children under 5 years of age, two-thirds of them are newborns, and there is still a significant gap between rural and urban areas [
16,
17]. In southwest of China, the neonatal mortality rate, infant mortality rate and under-5 mortality rate achieved a sustainable and substantial decrease from 2000 to 2014 after years of efforts [
17]. However, we can still find that 47.53% of children died at home and 34.53% of children didn’t even receive any medical care before death in southwest of China from a research analyzing the deaths of children under the age of five by verbal and social autopsy [
15]. Relevant reports are lacking in recent years on the improvement and current situation in southwest of China.
Research on danger signs in newborns has received less attention in China, especially in poor regions. To the best of our knowledge, there were no studies to assess the knowledge about neonatal danger signs in rual China. In this study, we chose mothers as the respondents because they were considered primary caregivers of newborns in local. The present study aimed to assess mothers’ overall level of knowledge of neonatal danger signs in Wenshan, a rural county of southwest of China, and to identify what danger signs were poorly recognized, and to explore associating social-economic factors with knowledge of neonatal danger signs.
Discussion
Reducing neonatal morbidity and mortality requires immediate caregivers’ recognition of neonatal danger signs and visits to the nearby health institution for medical care. In this study, 47 (42%) of the mothers had good knowledge, whereas 65 (58%) had poor knowledge, which is consistent with the report of Alex-Hart in Nigeria [
23], Degefa in southern Ethiopia [
6] and Guta in Dire Dawa of Ethiopia [
24]. Our results were also consistent with a systematic review and meta-analysis that reported that the overall pooled prevalence of mothers’ knowledge of neonatal danger signs was 40.7% [
25]. In this study, the proportion of knowledgeable mothers about danger signs was higher than the studies conducted in North West of Ethiopia (18.2%) [
26], Kenya (15.5%) [
27], 4 regions of Ethiopia (29.3%) [
28], Ambo town of central Ethiopia (20.3%) [
29] and northeast Ethiopia (28.2%) [
30]. But the proportion was lower than the findings of mothers attending public health institutions of Mekelle city [
31]. From this study, cultural variations among participants and differences in health care services delivery might have contributed to the differences in mothers’ knowledge about neonatal danger signs. Moreover, methodological differences, classification criteria differences, variations in time and study setting were the important reasons that contributed to the different results.
The two most commonly known danger signs were “Bluish or pale skin” (1.41 ± 0.65) and “chest indrawing” (1.41 ± 0.62). These are visible and obvious signs difficult to be missed. In a study reported by Sandberg in Uganda, “difficulty breathing” was recognized by a proportion (30%) of mothers by spontaneous response [
12]. “Bluish or pale skin” was the most common non-WHO-listed danger signs recognized by mothers in this study. An Indian research indicated that though not being recognized by WHO, some of the signs should be considered in danger signs list in developing countries as the signs are indicators of major causes of morbidity and mortality [
1].
In the present study, other signs such as “excessive crying” and “eyes draining pus”, “diarrhea”, and” yellow soles” were all poorly known. Signs of “not able to feed since birth, or stopped feeding well” got the lowest score among the 18 signs. In a study conducted by Khadduri et al., “difficulty feeding” was also identified as the least recognizable danger signs [
32]. The lack of awareness of “difficulty feeding” is the cause for concern in many situations because it is a condition life-threatening for newborns. In contrast to the findings of Ekwochi’s study, which mentioned “excessive crying” as a common non-WHO recognized neonatal danger signs [
7], we found this sign was less recognized than other signs in this study. These two signs could be unspecific signals rendering the mother unable to judge correctly whether the newborn is really sick. This difference may be due to mothers’ different perceptions of the severity of the problems. It can be assumed that poor knowledge of these danger signs may lead to delay in seeking appropriate medical care. Knowledge of danger signs is crucial to appropriate and timely care seeking behavior in undeveloped countries [
33,
34]. Educating should be strengthened on neonatal danger signs especially for items with poor scores.
Our study revealed that the number of ANC visits had a significant positive association with mothers’ knowledge about danger signs, which was consistent with the study conducted by Zaman in rural Bangladesh and Bayih in north-central Ethiopia [
35,
36]. This could be reasonably explained by the fact that those who attended ANC visits frequently are more likely to acquire the knowledge about danger signs. Community-based interventions integrating strategies of counseling and home visiting have been proved to reduce fetal and neonatal mortality [
37]. In this study, one-fourth of mothers attended the ANC visits less than 4, which shows they did not paying enough attention to prenatal examination. Efforts should be made to support pregnant mothers to attend ANC visits at the recommended frequencies.
Maternal age below 25 years old was found to be associated with increased odds of poor knowledge about neonatal danger signs. Similarly, the fact that younger mothers had lower knowledge of neonatal danger signs was also reported by another study conducted by Jemberia et al., which showed that caregivers older than 18 years old were 33% more aware of the neonatal danger signs than those younger than 18 [
22]. Another study found that all women aged < 18 years had poor knowledge about neonatal danger signs [
38]. In our study, 62% of mothers younger than 25 were primiparous, whom might be less experienced in caring for newborns. The youngest mother was only 15 years old while she was supposed to be getting married after reaching 20. The whole society should raise efforts to awake teenagers to avoid early marriage and early childbearing.
Ethnic minorities make up 59.6% of the population in the county. The ethnic minorities accumulated rich and valuable medical experience in the etiology, nomenclature, classification and treatment of diseases, building up strong cultural and religious beliefs [
39]. However, in this study, the mean score among Han mothers about neonatal danger signs was 22.76 ± 7.18 while that of ethnic minorities was only 15.76 ± 8.31. Mothers of ethnic minorities were 3.9 times more likely to possess poor knowledge about neonatal danger signs. This may be due to the fact that the ethnic minorities have a unique medical theory system. Ethnicity’s ancient primitive medicine culture was passed down through primitive myth and legend [
40]. The ethnic minorities live in a wide area with scattered villages and remote mountainous in southwest of China. Due to inconvenient transportation and poor information, ethnic minorities medicine has been passed down orally for a long time. In a qualitative study conducted by our team (yet to be published), we found ethnic minorities mothers self-medicated with herbs remedies for some diseases, such as neonatal jaundice, fever, diarrhea, etc. Herbal medicine is rich in the southwest of China, and the ethnic minorities people will take advantage of this unique natural resources by going up to the mountains to gather herbs. Cultural and religious factors have been reported as the main barriers influence caregivers’ perception of illness and care seeking behavior [
41,
42]. In addition, the low level of knowledge may be due to the fact that mothers of ethnic minorities generally married and gave birth before the age of 20.
Mothers with family financial difficulty were 4.9 times more likely to be unaware of danger signs than those mothers with no. This finding was consistent with a study performed in rural Bangladesh [
43]. Mothers in low-income settings may have limited recognition of danger signs. The consistency may be due to the fact that mothers who had financial difficulty were less likely to have higher education. On the contrary, mothers with financially secure are easier to get access to knowledge in different ways, such as mobile phones, the Internet and social network. A similar study reported by Anmut et al. [
44] that women with higher family income were 56% more knowledgeable about neonatal danger signs than those with low family income.
Previous studies confirmed that knowledge of neonatal danger signs was associated with caregivers’ educational level [
6,
30,
45]. Interestingly, educational background was not a significant factor in this study. An assumption might be the overall educational level of mothers was relatively low, which didn’t give educated mothers an advantage.
Strengths and limitations of the study
In this study, we used Likert 5 grading method and gave the corresponding score value, which makes us clearer how mothers perceive the different danger signs while some studies used unprompted answers to assess mothers’ recognition of the danger signs [
5,
29,
30,
44].
This study had several limitations. First, the respondents were mothers who had babies aged 0–12 months, thus the results may be affected by recall bias, and the mothers may fail to differentiate the period between the neonatal and postneonatal period. Second, though we had data quality control measures, the findings of our study might also be influenced by data collectors’ bias. Third, despite adjustments for socio-demographic factors, mothers’ healthcare service utilization related factors, due to the limitation of cross-sectional study design, there may be unidentified confounding variables. Fourth, there were three other towns that were not included in our study site, four health care centers may not be representative of all regions. The small sample size is also one of our limitations due to the difficulties in our process of data collecting in these areas. Besides, cluster effect may not be considered due to a limited sample size. In future studies, we plan to conduct more qualitative studies for further casual inference in more centers.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.