Associated ICU nurses' characteristics to clinical enteral nutrition knowledge at public hospitals in Sana'a, Yemen: A basis for remodeling safety and quality of care standards

Background: Nurses have a pivotal role in initiating and managing enteral nutrition (EN) and monitoring any potential complications. Yet, it is unclear whether Yemeni nurses have adequate knowledge to deliver and manage enteral nutrition safely. Therefore, the aim of this study was to assess the level of ICU nurses’ knowledge regarding the care of EN. Methods: A descriptive cross-sectional study was conducted from February 2019 to March 2019. A probability sampling method was used to recruit 174 nurses from four public hospitals in the capital city of Yemen. A self-administered 17-item questionnaire related to ICU nurses' knowledge regarding EN intervention was used to collect the data. Results: Based on our findings, only 10.9% of the respondents had an adequate level of knowledge about EN care, while 43.1% of them had a moderate level of knowledge, and 46.0% of them had an inadequate level of knowledge. Only 16.1% of the participants were knowledgeable of EN pre-administration care, whereas 5.80% of them were knowledgeable of EN administration care, and 9.20% were knowledgeable of EN post-administration care. There were significant associations between ICU nurses’ level of knowledge of EN care and their level of education and knowledge sources. Conclusions: The significant gap in ICU nurses’ knowledge regarding EN care identified implies the need to upgrade and refresh of the ICU nurses' knowledge by implementing a regular training program concerning EN care.


Introduction
Critically ill patients need regular enteral nutrition (EN) as an essential intervention to fulfil the body's dietary and physiological requirement 1 . EN is the recommended method of nutritional support for ICU ill patients who need extra nutritional calories because of the increased metabolic rate 2 .
It is considered as a medical therapy in law; hence, it should not be initiated only after ethical considerations have been made.
Previously, delivering of EN to critically ill patients was considered as a type of supportive care and not a therapeutic intervention 3 . Based on its benefits, it is more than a supportive therapy. It has been found that EN protects critically ill patients from malnutrition and subsequently improves body immunity and healing of tissues, which decreases their physiological stress effect, peptic ulcer, rates of infection 4,5 , the inflammatory response and maintains the function and integrity of the bowel 3,6 . Although maintaining balanced feeding via EN improves patients' health-related-outcomes, overfeeding leads to an increase in the mortality rate and must be avoided 7 .
Malnutrition among ICU ill patients is a universal public health concern, with a prevalence rate ranging from 40 to 60% 8 . According to previous studies, almost 70% of ICU patients acquired malnutrition during hospitalization 9 . To overcome this serious problem, EN should be initiated as early as possible for critically ill patients to avoid any risk of morbidity and mortality, which usually increases due to long patient hunger 10 . Previous studies and nutritional guideline have recommended that EN should be started within 24 to 48 hours of patients' admission to the ICUs or after the stability of the hemodynamic state, specifically after cardiac resuscitation 11 . In other words, once critically ill patients are admitted to ICUs, delivering appropriate nutrition becomes imperative 12 .
Despite the benefits of EN to critically ill patients, this process is usually associated with many complications, particularly if performed without an adequate level of knowledge 13,14 . Inappropriate delivering and poor care of EN

Amendments from Version 2
Title of the study: In response to the reviewer's comments, we amended and updated the study's title.

Discussion
In response to the reviewer's comments, we added a recommendation in the final paragraph to investigate the causal relationship, which might benefit in the design of future interventions focusing on specific factors.

Conclusion
In response to the reviewer's comments, we added to the concluding section a recommendation to use Vensim PLE's system dynamics modeling to demonstrate the impact of training on care improvements.

REVISED
can result into several complications, such as tube blockage or mal-positioning, nausea, vomiting, pulmonary aspiration, overfeeding, diarrhoea, treatment-related complications and delivery-site related complications like infection and agitation 15 . ICU nurses play a significant role in preventing such complications because they are responsible for identifying patients' nutritional risk, the calories needed, initiating and managing EN and monitoring any potential complications 16,17 . Accordingly, they must be knowledgeable enough to administer EN to critically ill patients safely 14,16 . To the best of the authors' knowledge, Yemeni nurses' enteral nutrition care knowledge has not yet been assessed, and they lack the necessary knowledge to safely deliver and care for enteral nutrition. Therefore, the aim of the present study was to explore Yemeni ICU nurses' knowledge regarding care of EN. Specifically, our study aimed to answer the following two research questions:

Design
A descriptive cross-sectional study design was used to assess the knowledge level of ICU nurses regarding EN care.

Setting
The study was conducted in ICUs at four public hospitals in Sana'a, the capital city of Yemen. These hospitals were selected as they are the referral hospitals for most people and the service fees are low.

Sample size determination
According to previous study, 71.1% of the respondents had a satisfactory total level of knowledge about EN 18 . Because our study was a cross-sectional, the following equation was used to calculate sample size: N=4pq/d 219 , where; N: required sample size, p: expected proportion of sample (from previous studies), q = 100 -p, and d= wanted precision (10% was taken in this study). Accordingly, a sample of 163 nurses was calculated. Assuming attrition of 10% = 16 + 163 = 179 nurses was required.
Sample and sampling method Stratified sampling was utilized to select the participants from the above-mentioned public hospitals. A list of the ICU nurses' name was obtained from each hospital. The required sample was drawn randomly from each list using a systematic random technique. All ICU nurses who had at least 6 months working experience, including both males and females with different educational qualifications, who were involved directly in ICU patients care, full-time employees and had agreed to participate were eligible to take part in this study. Nurses who included in the pilot study and those who were unwilling to participate were excluded. Based on these criteria, the eligible participants were approached by the researchers at their workplace. Out of 384 ICU nurses, 174 nurses were included in this study.

Study instrument
Based on previous related studies 10,20 , a self-administered questionnaire was adapted by the researchers. The questionnaire consists of 17 questions with four possible options to be answered. The final questionnaire (see Extended data 21 for a blank copy) is divided into two sections as follow: The first section is related to the socio-demographic characteristics of the participants: age, sex, level of education, working experience as a nurse, working experience as ICU nurse, training courses on EN and sources of knowledge about EN.
The second section is related to knowledge of ICU nurses regarding EN care and is divided into three subsections: 1. Knowledge of ICU nurses regarding before EN administration, which includes eight questions with a total of 32 responses.
2. Knowledge of ICU nurses regarding during EN administration, which involves four questions with a total of 16 responses.
3. Knowledge of the ICU nurses regarding after EN administration, which involves five questions with a total of 20 responses.

Scoring system
The 17 questions related to knowledge were assessed with "Yes" and "No" options. After correction of some reverse statements, a score of 1 was given for each correct response, while a score of 0 was given for each "incorrect" response. The maximum score for all correct answers was 68. Correct answers were calculated to obtain total scores for all questions of the three subsections. A score of 50% or less was considered inadequate, 51-75% moderate, while 76% and above was considered as adequate 22,23 .

Validity and internal consistency
Three experts in the EN from hospitals and Al-Razi University were invited to participate in examining the content validity for the instrument used in this study. Their comments concerning the tool accuracy, relevance, consistency, comprehensiveness and applicability for implementation were taken in consideration. A pilot study was conducted on 40 ICU nurses. Cronbach's alpha test was performed to examine the reliability of the questionnaire items. The result of the alpha was 0.78, which is acceptable.

Data collection
A self-administered questionnaire was distributed during the period of February to March 2019. The ICU nurses in the selected hospitals were invited to fill the questionnaire. Out of 179 questionnaires distributed, 174 were completed correctly and included in the final analysis.

Data analysis
The participants' responses were entered, cleaned, checked and explored using statistical software (IBMSPSS), version 22.0. The analyzed data was described using the mean values and standard deviations for continuous variables as well as the frequency and percentages for the categorical variables. Chi-square tests and Fisher exact test were conducted to find out the associations between the ICU nurses' knowledge and the selected socio-demographic variables. A p-value of ≤0.05 was reported as statistically significant.

Ethical consideration
Ethical clearance from the Ethics Committee of Al-Razi University was obtained for the current study. Then, an official written permission was also obtained from the managers of the selected hospitals prior to conducting the study. A written consent from all involved nurses was obtained prior to conducting the study.  Figure 2.

Participants' socio-demographic characteristics
Association between the ICU nurses' characteristics and their knowledge level of EN The current results indicated that there was a significant association between the level of education and sources of knowledge about EN and the overall level of participants' knowledge of EN care (P=0.011 and P=0.050, respectively). However, there were no significant associations between the participants' age, sex, experience as nurses, experience as ICU nurses and training courses and the overall level of knowledge regarding EN care (P= 0.291, 0.626, 0.453, 0.220 and 0.714, respectively). The detailed results of the association are presented in Table 2.

Discussion
The key finding in the current study was that the ICU nurses exhibited an inadequate level of knowledge concerning EN care. However, only 10.90% of the ICU nurses had an overall adequate level of knowledge regarding the EN care as a whole and the levels of adequate knowledge of it before administration, during administration and after administration were (16.10%, 5.80% and 9.20%, respectively). Such an inadequate knowledge among most Yemeni ICU nurses might be due to the inadequacy of in-service refreshing training' courses or because hospitals have not a clear and updated guideline protocol that ICU nurses can access and adhere to at work. As mentioned above, 59.20% of ICU nurses had never engaged in training courses about EN, and most had a three-year nursing diploma and had five-years working experience or less, which supports our speculation. Our findings highlighted the gap in ICU nurses' knowledge regarding the EN care, which indicates the necessity for conducting in-service training courses that focus on EN care. The results are consistent with an earlier study 10 which assessed knowledge and practice among 85 nurses working in ICU concerning EN care. The result revealed that ICU nurses had a low and inadequate level of knowledge about EN care. In another similar study 25 that assessed nurses' knowledge and practice regarding use of a nasogastric tube (NGT) in medications administration for ICU patients, an unsatisfactory level of knowledge among nurses was found. Additionally, another study 26 assessing nurses' knowledge level about nutrition revealed that nurses had poor knowledge of nutrition. However, the results of the current study are inconsistent with the result of Al-Hawaly, Ibrahim and Qalawa 18 who found that the majority of the respondents had a satisfactory overall level of knowledge concerning NGT nutrition administering. Likewise, the result disagrees with the result of Carlos, Costa and Simino 27 , who registered a satisfactory level of nurses' knowledge concerning nutritional therapy.
Another key finding of this study is that the level of ICU nurses' knowledge regarding EN care was significantly  Figure 1.
Overall level of the ICU nurses' knowledge regarding EN care Concerning the overall level of ICU nurses' knowledge on the different items of EN care, the results of the current study showed that the most (46.0%) of the participants had an  associated with the educational level (χ2 = 11.439, P= 0.012).
In other words, a high educational level in nursing was associated with a higher level of knowledge. This could be attributed to the fact that faculties of nursing focus more on the theoretical aspect, while health institutes pay considerable attention to the practical aspect. This finding is similar to

Joannes Paulus Tolentino Hernandez
Helene Fuld College of Nursing, New York, NY, USA First of all, reading the manuscript gives vital information of the quality of healthcare that warrants a quality overhaul. I am saddened by the findings although it is critical to make everyone know through research dissemination so readers can ruminate on culturally-relevant training and support for nurses, joint policy-making between implementers and surveillance agencies, and exercise of governance for reviewing the standards for patient safety.
Second, the descriptive statistics have underscored the reality that met the aims of the study. However, the data distribution could have been indicated and explored for skewness in order to expose the gaps in knowledge distribution. This knowledge distribution may reveal implicit 'disparity' (the access of resources and culturally inflated communicative actions). There is discourse of representation by the data points.
Third, the Chi-square tests and Fisher exact test did work, but if the authors can consider finding causal relationships then it would be a lot meaningful at the level of designing an intervention that targets specific factors/variables.
Fourth, establishing that Yemeni nurses have inadequate knowledge should not only be exploratory rather explanatory. Plans for training improvement must be articulated explicitly. It is clear that the authors just wanted to tell the world about a pervasive problem in Yemen, but the problem is at the healthcare system level. So, it may not only be a case of enteral nutrition.
Fifth, I have these suggestions: The title will be better if revised to: "Associated ICU nurses' characteristics to clinical enteral nutrition knowledge in Sana'a, Yemen public hospitals: A basis for remodeling safety and quality of care standards". ○ The discussion must connect the level of knowledge with (threat to) patient safety and quality improvements required. SWOT analysis is highly suggested.

○
On the other hand, please consider using system dynamics modeling at Vensim PLE. Here is a tutorial. This can be useful to show the impact of training to improvements in care. This serves as an attractive and alternative modeling. ○ p-value of less than .05 must be statistically significant.
○ predetermined study objectives. Based on the current findings, we stated that "this would enable decision-makers to develop a need-based plan to fill the existing gap, and Yemeni nurses' knowledge would allow to grow professionally and provide safe health-care services. However, we will consider your valuable comment in our future work.
COMMENT # 2: Second, the descriptive statistics have underscored the reality that met the aims of the study. However, the data distribution could have been indicated and explored for skewness in order to expose the gaps in knowledge distribution. This knowledge distribution may reveal implicit 'disparity' (the access of resources and culturally inflated communicative actions). There is discourse of representation by the data points.

RESPONSE:
Thank you for your comment. We conducted skewness in order to expose the gaps in knowledge distribution. As shown in table and figure, the distribution of knowledge was symmetric.

COMMENT # 3:
Third, the Chi-square tests and Fisher exact test did work, but if the authors can consider finding causal relationships then it would be a lot meaningful at the level of designing an intervention that targets specific factors/variables.

RESPONSE:
Thank you for your comment. We did, in fact, perform the Chi-square and Fisher exact tests. Unfortunately, because it was not one of our study aims, we did not explore the causal relationships. However, we have included this useful suggestion in our study recommendations for further research. Please see lines 4-6 in the last paragraph of the discussion section.
COMMENT # 4: Fourth, establishing that Yemeni nurses have inadequate knowledge should not only be exploratory rather explanatory. Plans for training improvement must be articulated explicitly. It is clear that the authors just wanted to tell the world about a pervasive problem in Yemen, but the problem is at the healthcare system level. So, it may not only be a case of enteral nutrition.

RESPONSE:
Thank you for your insightful observation and comment. In fact, our study objective was to explore the level of ICU Nurses' knowledge regarding enteral nutrition. Based on our study findings, we have discussed and explained the findings, in discussion section. Please see the discussion part. Based on your advice, we have already added a recommendation for additional research to examine the causal relationships.
COMMENT # 5: Fifth, I have these suggestions:

COMMENT:
The title will be better if revised to: "Associated ICU nurses' characteristics to clinical enteral nutrition knowledge in Sana'a, Yemen public hospitals: A basis for remodeling safety and quality of care standards". RESPONSE: Thank you for your comment.
We have updated the study title based on your suggestion. COMMENT: The discussion must connect the level of knowledge with (threat to) patient safety and quality improvements required. SWOT analysis is highly suggested.

RESPONSE:
Thank you for your comment. We employed the elements of discussion in this present study. We humbly believe that SWOT analysis in the discussion part can be used in our future interrogations. COMMENT: On the other hand, please consider using system dynamics modeling at Vensim PLE. Here is a tutorial. This can be useful to show the impact of training to improvements in care. This serves as an attractive and alternative modeling. RESPONSE: Thank you for your comment and useful tutorial link. Indeed, our study was not aimed to show the impact of training to improvements in care. Nevertheless, we will take this into consideration in our future research endeavor and will put this in one of our conclusion and recommendations. Please see lines 5&6 of the conclusion section. COMMENT: p-value of less than .05 must be statistically significant.
be carefully selected.
Methods: Please clearly describe how to perform the multinomial logistic regression. How to obtain the p-values in Table 2? Multinomial logistic regression? Chi-square test?

4.
Methods: Because the cells in Table 2 contained under <5, Fisher exact test is more suitable than chi-square test.

5.
Discussion: If the authors used the different questionnaire used in the previous study, how and why the authors could compare the results? 6.
Conclusions: Is there any evidence that training courses in EN management improve clinical outcomes? Based on the results of your study, the only thing the authors revealed was that low SES was associated with inadequate knowledge. Experience and training courses were not associated with improved knowledge. Therefore, the conclusion should be based on their results.

7.
Discussion: What is new in this study? What can the readers act from this paper all over the world? 8.

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Partly Are the conclusions drawn adequately supported by the results? Partly COMMENT # 5: Methods: Because the cells in Table 2 contained under <5, Fisher exact test is more suitable than chi-square test.

RESPONSE:
Thank you for your comment. We fully agree with you. We use the Fisher exact test to re-analyze the data in this table and re-write the table results accordingly. Please see  Table 2. Please also see line 4 for the addition of this test to the data analysis section.
COMMENT # 6: Discussion: If the authors used the different questionnaire used in the previous study, how and why the authors could compare the results?

RESPONSE:
Thank you for your comment. Our questionnaire was not significantly different from those used in previous studies. We simply modified an existing questionnaire rather than creating a new one. As a result, we compared our study findings to those of related studies with similar intended outcomes.

COMMENT # 7:
Conclusions: Is there any evidence that training courses in EN management improve clinical outcomes? Based on the results of your study, the only thing the authors revealed was that low SES was associated with inadequate knowledge. Experience and training courses were not associated with improved knowledge. Therefore, the conclusion should be based on their results.

RESPONSE:
Thank you for your comment. Thank you for your comment. Our conclusion is derived from our results and there is no sentence implied that training courses in EN management improve clinical outcomes but improve the awareness and knowledge. this is what we concluded based on our findings and this is supported by many previous studies. Please see our conclusion section.
COMMENT # 8: Discussion: What is new in this study? What can the readers act from this paper all over the world?

RESPONSE:
Thank you for your comment. This study, like many others around the world, has clear objectives. These are stated clearly at the end of the introduction section. These are explicitly stated at the end of the introduction. According to the study findings, we believe that the situation regarding Yemeni nurses' enteral nutrition care knowledge becomes clear. This would help decision makers develop a plan to fill this knowledge gap, and as a result, Yemeni nurses' knowledge would grow professionally, and they would be able to provide safe health care services. Furthermore, the findings of this study will serve as a data base for future research and will provide potential readers with information about the current situation of Yemen's nurses. We've added a paragraph to explain the study's implications. Please see the discussion section, paragraph 4, Lines 1-6. https://doi.org/10.5256/f1000research.27625.r67783 In results, the use of some terms (such as most, majority) in describing the results need a revision. For example: 49.40% is not a majority.
○ Table 2 presents the frequency and percentage of the participants and compares between the subgroups using the Chi-square tests. I think it will be better if the table presents the knowledge means scores and standard deviations of the subgroups and compare between them by the t or f tests. After that multiple regression can be performed.

Discussion:
Discussion must be on the main important findings based on the objectives of the study. No need to write two paragraphs at the beginning of the discussion about the participants' characteristics such as age, gender, educational levels, and work experience. Describing those characteristic in results part is enough.

Gamil Ghaleb Alrubaiee
COMMENT # 1: Abstract: The results part in the abstract should focus on the main results such as the areas with high level of knowledge and areas with lower level of knowledge.
Remove the excessive description of the sample socio-demographic characteristic.

RESPONSE:
Thank you for your comment. The results section has been rewritten based on your recommendation. Please see the abstract's results section.

COMMENT # 2:
Conclusion part has been written as a recommendation only. It will be better if you summarize the findings then make a brief recommendation.

RESPONSE:
Thank you for your comment. Indeed, the findings were summarized in the results section, and the conclusion section focused on the main result, the knowledge gap, and the recommendations based on that result.

COMMENT # 3:
Methods: Under the setting subtitle no need to mention the names of the hospitals. Maintaining the confidentiality of participant hospitals. No need to mention the duration of the study under the setting subtitle, it is mentioned under data collection.

RESPONSE:
Thank you for your comment. We have followed your recommendations.

RESPONSE:
Thank you for your comment. We have followed your recommendations.

COMMENT # 5:
Instead of citing the tool they have developed and used, authors must mention and cite the references and literature that had been used in developing the tool.

RESPONSE:
Thank you for your comment. The references for adapted tool were cited as recommended.
COMMENT # 6: Per authors, the second part of data collection instrument contains three subsections which are the before EN administration, during EN administration, and after EN administration. Accepting that this categorization is correct, it is not clear which items are under each subsection. Moreover, most of the 17 items are not related to the before, during, and after EN management, they are related to general knowledge about the EN (such as the indications, goals, benefits, complications, routs, methods, types and contents of EN formula, ........ etc.), even there is an item asking about the types of nutritional support in general (enteral, parenteral, combination, ..etc.). More suitable grouping and categorizing must be used to organize the instrument or to leave it without any categorization it will be better. Instead of citing the tool they have developed and used, authors must mention and cite the references and literature that had been used in developing the tool.

RESPONSE:
Thank you for your comment. The items under each subsection were mentioned to be clear for the potential readers. We also replace the word "management" by