Reduction of Hospital Acquired Infections. Hand Hygiene among Nursing Students in Zambia

CONTENTS

DECLARATION

ABSTRACT

ACKNOWLEDGEMENTS

LIST OF ABBREVIATIONS

GLOSSARY

LIST OF TABLES

LIST OF FIGURES

LIST OF EQUATIONS

CHAPTER 1 INTRODUCTION
1.1 PREVALENCE OF HOSPITAL ACQUIRED INFECTIONS (HAI) IN DEVELOPING COUNTRIES
1.2 EPIDEMIOLOGIC TRIAD OF HAI
1.3 PRIMARY PREVENTION OF HAI THROUGH EFFECTIVE HAND HYGIENE
1.4 SIGNIFICANCE OF THIS STUDY

CHAPTER 2 LITERATURE REVIEW
2.1 LITERATURE SEARCH STRATEGY
2.2 SUMMARY OF LITERATURE SEARCH
2.2.1 Effectiveness of hand hygiene
2.2.2 Hand hygiene compliance in lower middle-income countries (LMIC)
2.2.3 Demographics and training factors associated with hand hygiene compliance in LMIC
2.2.4 Knowledge, attitude and practices associated with hand hygiene in LMIC
2.2.5 Hand hygiene compliance in Zambia
2.3 RESEARCH GAPS

CHAPTER 3 METHODS
3.1 BACKGROUND
3.2 AIM
3.3 OBJECTIVES
3.4 EPISTEMOLOGICAL APPROACH
3.5 STUDY DESIGN
3.6 SETTING
3.7 SAMPLING APPROACH/FRAME
3.8 SAMPLE SIZE
3.9 ETHICS
3.10 INCLUSION AND EXCLUSION CRITERIA
3.11 RECRUITMENT AND DATA COLLECTION METHODS
3.12 MINIMISING BIAS
3.13 POTENTIAL CONFOUNDERS
3.14 INSTRUMENT
3.15 DATA RESPONSES TO HAND HYGIENE
3.15.1 Knowledge related questions
3.15.2 Perception related questions
3.16 PILOT TESTING
3.17 DATA CLEANING AND ANALYSIS
3.17.1 Descriptive statistics
3.17.2 Inferential statistics - Bivariate Analysis
3.17.3 Inferential statistics - Multivariate Analysis

CHAPTER 4 RESULTS
4.1 DESCRIPTIVE INFORMATION
4.1.1 Demographic and training factors for nursing students
4.1.2 Hand hygiene Knowledge of Participants
4.4.3 Hand hygiene Perception of Participants
4.2 ASSOCIATION BETWEEN DEMOGRAPHICS/ TRAINING FACTORS AND HAND HYGIENE KNOWLEDGE - CROSSTABULATION AND FISHER'S EXACT TEST
4.2.1 Year of Study
4.2.2 Program of Study
4.2.3 Routine use of Alcohol-based hand rub
4.2.4 Average percentage of patients to develop HAI
4.3 UNIVARIATE ANALYSIS WITH MULTINOMIAL REGRESSION
4.3.1 Selection of predictor variables at p=0.1 cutoff
4.4 MULTIVARIATE ANALYSIS WITH MULTINOMIAL REGRESSION
4.4.1 Adjusted odds ratio for nursing students hand hygiene knowledge by demographic and training factors
4.5 MISSING DATA

CHAPTER 5 DISCUSSION
5.1 PUBLIC HEALTH CONTEXT
5.2 IMPLICATIONS OF THIS STUDY'S FINDINGS
5.2.1 Knowledge of hand hygiene
5.2.2 Perception about hand hygiene
5.2.3 Program of Study
5.2.4 Year of Study
5.2.5 Routine use of alcohol-based hand rub
5.3 RESEARCH PROCESS
5.3.1 Critical analysis of the research process
5.3.2 Study Strengths
5.3.3 Study Limitations
5.3.4 Validity and Reliability
5.3.5 Generalisability
5.3.6 Lessons learnt
5.3.7 Why Chi-Square was not used
5.4 RECOMMENDATIONS

CHAPTER 6 CONCLUSION

REFERENCES

APPENDIX 1 - DATA COLLECTION SHEET

APPENDIX 2 - INFORMATION SHEET AND CONSENT FORM

APPENDIX 3 - LOCAL ETHICAL APPROVAL FROM TDRC

APPENDIX 4 - GATE KEEPER PERMISSION GRANTED

APPENDIX 5 - ETHICAL APPROVAL FROM UNIVERSITY

APPENDIX 6 - APPROVED PROPOSAL OUTLINE

APPENDIX 7 - HAND HYGIENE TECHNIQUE USING ALCOHOL

APPENDIX 8 - HAND HYGIENE TECHNIQUE USING SOAP AND WATER

APPENDIX 9 - MISSING DATA

DECLARATION

No portion of this work has been submitted in support of an application for degree or qualification of this or any other university or institution of learning

Imukusi Mutanekelwa

March 2019

ABSTRACT

Background

The low compliance to effective hand hygiene has continued to fuel the high prevalence of Hospital Acquired Infections (HAI) in Africa. The large number of nursing students has a potentially high impact at reducing the HAI public health problem in Zambia, however, there is insufficient information on the nursing student's demographic/training factors affecting their hand hygiene knowledge which reduces the HAI risk.

Objective

To determine the demographic/training factors associated with nursing student's hand hygiene knowledge in Solwezi, Zambia thus provide information for action necessary to reduce HAI.

Methods

A quantitative cross-sectional survey using primary data collected via a WHO validated self-administered questionnaire distributed to student nurses > 18 years at Solwezi College of Nursing. 167/206 (81.1%) participants were recruited via stratified random sampling. Descriptive statistics highlighted potential relationships between demographic/training factors and students hand hygiene knowledge which reduces HAI; potential relationships were analysed using Fisher's exact test and multinomial logistic regression.

Results

Most (60.5%) nursing students had moderate hand hygiene knowledge. Using Fisher's exact test, hand hygiene knowledge was significantly associated with three training factors: year of study (p=0.018), program of study (p=0.003), routine use of alcohol-based hand rub (p=0.017), and one perception factor: average percentage of hospitalised patients who develop HAI (p=0.015). Regression analysis showed that only program of study was significantly associated with hand hygiene knowledge; general nursing students were 24 times more likely to have a moderate knowledge score compared to public health nursing students, adjusted odds ratio = 24.859, p = 0.029.

Conclusion

Public health nursing students in Solwezi, Zambia posed the highest risk of spreading HAI owing to inadequate hand hygiene knowledge, hence tailor made interventions (e.g. refresher lessons) should consider the different program specific attributes as guided by this study.

Keywords

Nurse, infection prevention, hand hygiene, Hospital Acquired Infection, primary prevention, Zambia

Abstract word count: 300

Dissertation word count: 10 765 (excluding cover page, declaration, abstract, acknowledgement, abbreviations, glossary, contents, list of tables and figures, tables and charts, references, and appendix)

ACKNOWLEDGEMENTS

I want to thank my enrollment advisor Guilherme Degasperi for the wonderful job of guiding me at the start of the MPH program. I want to sincerely thank my dissertation advisor, for her invaluable support, strong positive criticism of my work and guidance throughout the research process. I am also grateful to all the module lecturers and classmates, student support manager Ruchi Bansal and Gillian Woolhead for the assistance provided throughout the MPH course.

I would also like to thank the entire management for Solwezi College of Nursing and student nurses for the support I received.

I also sincerely thank my parents Mr and Mrs Mutanekelwa for the motivation and encouragement throughout the MPH journey. Finally, I also thank my loving wife Chanda Mulenga-Mutanekelwa for the heartfelt encouragement and support during the MPH journey.

LIST OF ABBREVIATIONS

Abbildung in dieser Leseprobe nicht enthalten

GLOSSARY

Hand hygiene: as a general term applies to “performing handwashing, antiseptic hand wash, alcohol-based hand rub, surgical hand hygiene/antisepsis” for the purpose of removing soil or microorganisms in order to prevent cross-contamination and minimise nosocomial infections (WHO, 2009).

- Alcohol-based handrub formulation: “ an alcohol-containing preparation (liquid, gel or foam) designed for application to the hands to kill germs” (Wretling, 2014).
- Handrubbing: treatment of hands with an antiseptic handrub (alcohol-based formulation) (Wretling, 2014).
- Handwashing: “ washing hands with plain or antimicrobial soap and water” (Wretling, 2014).
- Surgical hand hygiene/antisepsis: “handwashing or using an alcohol-based handrub before operations by a surgical personnel” (WHO, 2009).

HAI (also known as nosocomial infections) - infections that occur during the care of a patient in the hospital/healthcare facility which was absent or in the incubating period at admission (WHO, 2009).

Knowledge - “facts, information, and skills acquired through experience or education; the theoretical or practical understanding of a subject” (Cambridge Dictionary, 2018).

Perception - “a belief or opinion, often held by many people and based on how things seem” (Cambridge Dictionary, 2018).

LIST OF TABLES

Table 1 shows factors in the hospital environment with the potential to contribute to HAI

Table 2 shows demographic and training characteristics of participants

Table 3 shows the student responses which were aggregated to contribute to the overall knowledge score (i.e. dependent variable)

Table 4 shows cross tabulation between year of study and hand nursing student's hygiene knowledge

Table 5 shows Fisher's exact test between year of study and hand nursing student's hygiene knowledge

Table 6a shows cross tabulation between program of study and nursing student's hand hygiene knowledge

Table 7b shows Fisher's exact test between program of study and nursing student's hand hygiene knowledge

Table 8 shows cross tabulation between 'routine use of alcohol hand rub' and nursing student's hand hygiene knowledge

Table 9 shows Fisher's exact test between 'routine use of alcohol hand rub' and nursing student's hand hygiene knowledge

Table 10 shows cross tabulation between nursing student's hand hygiene knowledge and perception of average % of patients who develop HAI

Table 11 shows Fisher's exact test between nursing student's hand hygiene knowledge and perception of average % of patients who develop HAI

Table 12 shows variables which were univariately associated with students hand hygiene knowledge at p value 0.1 as cut off. Also shown are test of two assumptions using VIF and Durbin Watson test

Table 13 shows the results of multivariate multinomial regression results, the reference group was poor knowledge score. Furthermore, SPSS used the last category of each predictor variable as the reference hence it's Exp (B) or p is denoted 0b

Table 14 shows the testing of chi-square assumptions for this study, a stratified random sample of nursing students was used

LIST OF FIGURES

Figure 1 shows the 5 moments when to perform hand hygiene to reduce HAI's

Figure 2 shows the winnowing process used for the selected articles reviewed

Figure 3 shows North Western Province (on the small map of Zambia) and location of Solwezi district

Figure 4 shows nursing students knowledge score, 35.3% had poor score, 60.5% had moderate, 4.2% had a good score. The median knowledge for all participants was 10.00 with IQR

Figure 5 shows participants opinion about the average percentage of hospitalised patients who will develop HAI (between 0 and 100%)

Figure 6 shows students opinion of the effectiveness of hand hygiene in preventing HAI (with 0 as ‘Very Low', 1 as ‘Low', 2 as ‘High' and 3 as ‘Very High'

Figure 7 Students perception of the importance of hand hygiene at SCN (with 0 as ‘Very Low Priority', 1 as ‘Low priority', 2 as ‘High Priority' and 3 as ‘Very High Priority')

Figure 8 shows students opinion of the percentage of situations requiring hand hygiene which HCWs at SGH actually perform either by hand rubbing or hand washing (between 0 and 100%)

Figure 9 Students perception of the effort required to perform good hand hygiene when caring for patients (with 0 meaning ‘No Effort' and 6 for ‘Big Effort')

Figure 10 shows students opinion of the percentage of situations requiring hand hygiene which students actually perform either by hand rubbing or hand washing (between 0 and 100%)

LIST OF EQUATIONS

Equation 1 shows the regression model equation for predicting a student's knowledge score, n = 6 predictor variables

CHAPTER 1 INTRODUCTION

1.1 PREVALENCE OF HOSPITAL ACQUIRED INFECTIONS (HAI) IN DEVELOPING COUNTRIES

World Health Organisation (WHO) defines Hospital Acquired Infections (HAI) as infections that occur during the care of a patient in the hospital/healthcare facility which was absent or in the incubating period at admission (WHO, 2009).Because most (84.4%, 124/147) developing countries (Zambia inclusive) don't have a functional national surveillance system in place, the prevalence of HAI is either unknown or underestimated because of the intricacies of making such a diagnosis (WHO, 2010; Nejad et al., 2011). Nonetheless,HAI is a major public health problem with the prevalence ranging from 5.7% - 19.1% and up to 37% for those admitted in intensive care units in developing countries (WHO, 2009; Nejad, et al., 2011; Khan, et al., 2017). Despite limited evidence on the economic burden of HAI in Zambia, in developed nations, the annual direct medical cost is more than US$ 35 billion due to prolonged stay in hospital which attracts more laboratory investigations, treatments and nursing care (WHO, 2016). An understanding of how HAI spreads is necessary to improve existing public health interventions that reduce its prevalence, this is discussed below.

1.2 EPIDEMIOLOGIC TRIAD OF HAI

Agent (or Disease)

The agent is simply an infectious pathogen or microorganism and must be usually present for HAI to occur, other factors include pathogenicity and dose (Gordis, 2014). Microorganisms colonising the hands are of two types, resident or transient flora. Healthcare workers (HCW) and students often acquire transient flora after contact with contaminated surfaces in the patient's environmentor patients. Transient flora is present on the skins superficial layers and thus more likely to be removed by hand hygiene(WHO, 2009).

Host factors

The host refers to the person able to acquire HAI. Host factors increasing the probability of disease include old age, immunosuppression, vaccination status, the presence of underlying disease, hygiene compliance (Syrjanen, 2014).

Environment

The environment considers factors which can increase the probability of exposure and these include factors associated with HCW, patients, and vaccinations(Syrjanen, 2014).

Abbildung in dieser Leseprobe nicht enthalten

Table 1 shows factors in the hospital environment with the potential to contribute to HAI

Source: (Syrjanen, 2014, pp. 1478) In summary, a disease agent can potentially cause HAI in a susceptible host present in the hospital environment. Despite various factors mentioned (see Table 1), this research focused on the HCW associated factors causing HAI with emphasison handhygiene, this is further discussed below.

1.3 PRIMARY PREVENTION OF HAI THROUGH EFFECTIVE HAND HYGIENE

Effective hand hygiene must be employed by all HCW's including nurses to reduce the HAI incidence (i.e.primary prevention). Effective hand hygiene includes knowing the right moments when to wash hands during routine clinical practice (five moments of hand hygiene), how to wash hands during those moments (WHO eleven steps), and what to use to wash hands (alcohol-based hand rub,soap). These five momentsshown in (WHO, 2009, pp. 123)

Figure 1 include: “before and after touching apatient, before anyaseptic procedure, after exposure to bodilyfluids and lastly after touching patientsurroundings” (WHO, 2009, pp. 123)

Abbildung in dieser Leseprobe nicht enthalten

Figure 1 shows the 5 moments when to perform hand hygiene to reduce HAI's

Source: (WHO, 2009, pp. 123)

Awareness of the five moments to practice hand hygiene is not used independently to prevent HAI without knowledge of the hand hygiene technique to use based on different antiseptics available. When alcohol-based hand rub is used, different areas of the hands must be ‘rubbed' using eight steps for a recommended duration of 20 - 30 seconds (see APPENDIX 7 - hand hygiene technique using alcohol). In contrast when using soap and water the hand washing steps increases to 11 and duration to 40 - 60 seconds (see APPENDIX 8 - HAND HYGiENE TEcHNiQuE uSiNG SoAP AND WATER) (WHO, 2009).

1.4 SIGNIFICANCE OF THIS STUDY

As mentioned above, despite HAI being preventable through effective hand hygiene to interrupt disease transmission, there is still a high HAI prevalence and low hand hygiene compliance rate. Poor hand hygiene can cause catheter-associated infections, surgical site infections, ventilator-associated pneumonia. A Sub-Saharan study by Chu et al., (2015) showed that the prevalence of post caesarian section surgical site infections was 7% and increased the median length of hospital stay from 7 days to 21 days (Chu et al., 2015). In a resource-limited setting like Zambia, treatment of HAI's (i.e. secondary prevention) is an economic burden due to costs related to prolonged hospital stay, patients out of pocket, increased chances of drug resistance, post-discharge complications, extra diagnostic and medical procedures (De Angelis et al., 2010). Nurses can help to reduce HAI disease burden because they are the largest (47.2%) in number and most widely distributed among all HCW's according to Zambia's National Human Resources for Health Strategic Plan 2011-2015. Despite the inequitable distribution of nurses and other HCW's, strengthening measures to improve hand hygiene among nursing students would potentially have a positive impact on reducing HAI (GRZ/MOH, 2011).

This study examines the public health aspect of the problem of HAI by focusing on primary prevention using student nurses rather than secondary prevention. This was done by exploring the specific demographics and knowledge affecting nursing students hand hygiene practice aimed at reducing HAI in Solwezi, Zambia. Furthermore, this is consistent with primary prevention efforts being currently spearheaded by the Zambian Ministry of Health in line with the 2016 health reforms (ZNPHI, 2017). Hence the information generated from this research will inform policymakers at General Nursing Council of Zambia (GNC) and contribute to initiatives aimed at improving students hand hygiene techniques during clinical placements in hospitalsto reduce the HAI public health problem.

CHAPTER 2 LITERATURE REVIEW

This chapter's objective was to review the literature related to nursing students demographic and training factors associated with handhygiene knowledge, attitude and practices (KAP) that reduce HAI using the subheadings:

a) The effectiveness of hand hygiene to avert HAI
b) Hand hygiene compliance in lower middle-income countries (LMIC)
c) Demographics and training factors associated with hand hygiene among nurses in LMIC
d) KAPassociated with hand hygiene among nurses in LMIC
e) Hand hygiene compliance in Zambia

2.1 LITERATURE SEARCH STRATEGY

Various publications including dissertations related to nursing students hand hygiene KAP and its association to demographic and training factors were retrieved. Academic health databases electronically searched included Google scholar, University of Liverpool online library, Cochrane, CINAHL, PubMed, Zambia's Ministry of Health and University of Zambia library. International organisations dealing with hand hygiene were also searched such as WHO, UNICEF. To improve relevance and narrow down on search findings, the following keywords with synonyms and Boolean operator were used “hand washing compliance OR hand hygiene AND factors OR demographics OR barriers AND nurse OR healthcare worker AND Nosocomial infections OR Healthcare associated infections OR Hospital-acquired infection”

The following filters were applied in the database search engines;

a) Time filter: 10 years (2008 - 2018) for Zambian studies, 5 years (2014 - 2018) for other studies in LMIC
b) Language filter: English

Given that the burden of HAI is larger in LMIC's compared to developed nations with a prevalence of up to 5.7% - 19.1% as mentioned earlier (Nejad et al., 2011) and that LMIC's share similar healthcare challenges regarding IPP (Vilar-Compte et al., 2017), hence the literature search focused on LMIC's.

The resulting articles after filtering in the search engines were subjected to first level screening via title and abstract. Potential articles with relevant keywords (as shown above) were scrutinised using the inclusion criteria below;

a) Country: LMIC's (including Zambia) as defined by World Bank as having GDP between $1046 - $4125 with the list of countries obtained from the World Bank website (World Bank, 2018)
b) Population: Nurses included in the study
c) Study type: Quantitative study (see Figure 2).

All the downloaded full articles of the selected abstracts were subjected to a second level screening which involved critical appraisal of each article to determine if it would be included in this study using the following minimum criteria;

a) Presence of a clearly focused research question
b) Use of valid methods to tackle research questions
c) Importance of study results
d) Applicability of study findings to Zambia

Abbildung in dieser Leseprobe nicht enthalten

Figure 2 shows the winnowing process used for the selected articles reviewed

Source: (PRISMA, 2009)

Title screening yielded 337 abstracts for previewing. Following abstract screening, full articles for 75 studies were downloaded and critically appraised using standardised checklists dependent on study type. Finally, only 31 selected articles were included in the review (see Figure 2). Apart from three Zambian studies, the rest were conducted in West Africa, Sub-Saharan Africa, North Africa, the Middle East and South Asia while non-LMIC were excluded in the literature review.

2.2 SUMMARY OF LITERATURE SEARCH

2.2.1 Effectiveness of hand hygiene

Evidence shows hand hygiene as the most effective intervention to prevent HAIs, thus WHO recommends it as the most important measure (Labrague, et al., 2017, pp.1). Effectiveness against removing different microorganisms has been demonstrated by various investigators using soap, alcohol-based solutions, 4% chlorhexidine gluconate solutions, povidone iodine in Brazil, Tanzania, Pakistan, and North America and Europe (Ojajarvi , 1980; Luby, et al., 2001; Hernandes, et al., 2004; Bloomfield, et al., 2007; Pickering, et al., 2010).

2.2.2 Hand hygiene compliance in lower middle-income countries (LMIC)

Despite the effectiveness, various cross-sectional studies and one cluster randomised controlled trial in LMIC have consistently shown a low compliance towards the WHO five moments for hand hygiene with 14.6%(Rynga et al., 2017), 16.5% (Abdella et al., 2014), 19.5% (Santosaningsih et al., 2017) and 22.0% (Kolola and Gezahegn, 2017). Reasons for the low compliance include high workload, differences in understanding and availability of hand hygiene resources (Chitimwango, 2017; Musu, et al., 2017; Nohemi and Graciela, 2017). All the studies utilised the WHO direct observation checklist and all except one study (Kolola and Gezahegn, 2017) did not use the WHO five moments for hand hygiene standardised questionnaires on knowledge and perception of HCW. Half of the studies sampled all HCW and the other studies used a random sample of 284 (Santosaningsih et al., 2017) and 405 (Abdella et al., 2014). A major limitation of the quantitative studies included Hawthorne effect where participants changed behaviour because of being observed performing hand hygiene. In comparison, this study did not directly observe student nurses for two reasons; firstly the aim was not to determine hand hygiene compliance and secondly differences in data collection methods hence the Hawthorne effect limitation did not apply to this study. Another common restriction for the cross-sectional studies was the failure to establish a temporal relationship that having adequate hand hygiene knowledge and good attitude translated to good hand hygiene practice leading to reduced HAI through improved compliance. This can be overcome by using more robust experimental studies by providing extensive hand hygiene classroom and practical training to nursing students after that measuring the effect on hand hygiene KAP and compliance, however, this was beyond the researcher's ability because of limited time and resources.

Based on the literature findings healthcare facilities and nursing schools were able to adopt tailored hand hygiene interventions to improve compliance and reduce HAI through refresher training and provision of alcohol hand rub (Abdella et al., 2014). Zambia can benefit from carrying out a similar study to assess nursing student's awareness of hand hygiene. Hence this study assessed the differences in nursing students understanding of hand hygiene separated according to demographics and training factors using a WHO validated questionnaire via a cross-sectional design.

2.2.3 Demographics and training factors associated with hand hygiene compliance in LMIC

Numerous cross-sectional studies in LMIC have consistently shown that knowledge about IPP such as hand hygiene was significantly associated with education in Ghana, Palestine and Ethiopia (Ocran and Tagoe, 2014; Fashafsheh et al., 2015; Desta et al., 2018), one Ethiopian study contradicted the finding (Hussen et al., 2017). Similarly, gender has been associated with different knowledge levels towards IPP such as hand hygiene in Palestine and Ethiopia (Fashafsheh et al., 2015; Hussen et al., 2017), while other authors have found no such association in India (Diwan et al., 2016). Authors Desta and colleagues (2018), showed that knowledge of IPP was better with old age while Langoya and Fuller, (2015) in South Sudan, contradicted by reporting it was better in the younger age group. In contrast, other authors found no association between knowledge and age (Fashafsheh et al., 2015). Equally, the proportion of HCWs knowledge of IPP was better with longer work experience in Ethiopia and India (Desta et al., 2018; Ojah and Das, 2018) while other authors found no relationship (Fashafsheh et al., 2015; Hussen et al., 2017). Only one study has shown that the department where HCWs were placed was associated with IPP knowledge (Hussen et al., 2017). Regardless of any previous formal training in IPP, there was no association with knowledge (Fashafsheh et al., 2015; Langoya and Fuller, 2015; Diwan et al., 2016), this was not the case for a study by Hussen et al., (2017) were an association was determined. The disparities among studies in demographics affecting hand hygiene are likely due to differences in selected settings (entire hospital versus Intensive Care Unit), participant characteristics (nurses versus all HCWs) and sampling methods, differences in statistical methods and cutoff used for regression. The Common limitation themes for the studies reviewed include the inability to establish a temporal relationship, Hawthorne effect for observation checklist, and social desirability bias for self-administered questionnaires. Social desirability occurred in the Ethiopian study by Desta et al., (2018) because of potentially sensitive questions used in the modified questionnaire neither piloted nor previously validated and sourced from CDC, other hospitals and literature. This potential bias was reduced in this study by piloting and using a WHO validated tool. The limitations caused by Hawthorne effect and temporal relationship were discussed earlier in section 2.2.2. As consistently shown throughout history from the time of John Snow, understanding the epidemiologic profile of persons spreading HAI (i.e. student nurses) has an impact on reducing disease incidence. Other studies have identified specific HCW at high risk of spreading HAI, and developed tailored interventions (prioritising specific HCW) that successfully reduced HAI (Abdella et al., 2014; Kolola and Gezahegn, 2017). Zambia can benefit from adopting similar approaches hence this study mapped the epidemiologic profile of student nurses based on demographic and training factors affecting hand hygiene knowledge and indirectly the risk of HAI.

2.2.4 Knowledge, attitude and practices associated with hand hygiene in LMIC

Descriptive statistics of IPP such as hand hygiene reported that HCWs inclusive of nurses had a moderate knowledge score of 75.2% using a three-scale scoring system of good (>75%), moderate (50 - 74%), and poor (<50%) (Kudavidnange et al., 2014). The former study informed this study because of employing the same scoring system and WHO data collection tool. The study by Olalekan et al., (2018) reported that nurses knowledge score was mostly good (72.2%), very good (13.0%) and poor (14.8%). The knowledge score ranged from 0 - 16 with scoring 0-2 meaning ‘no', 3-5 poor, 6-8 fair, 9-11 good, 12-14 very good, >15 excellent (Olalekan et al., 2018). Another study showed that 85.1% of HCW had moderate knowledge about hand hygiene with no HCW having high knowledge. The study also used a three-scale scoring system with a total score of 25, with >75% meaning ‘good', 50 - 74% moderate and <50% poor (Salman et al., 2018). The study by Lien et al., (2018) showed that HCWs in urban and rural settings showed no significant differences in knowledge profile. Another study measured knowledge score on a two- point scale of good and poor and showed that most HCWs had a good knowledge (99.3%) and a few with poor knowledge (0.7%) (Hussen et al., 2017).

Taking into consideration that Zambian student nurses are taught for eleven weeks about infection prevention practices (IPP) with emphasis to hand hygiene in the first year foundation courses, it was necessary to gauge student's knowledge and perception of hand hygiene aimed at reducing HAI and also identify any demographic/training factors affecting student's hand hygiene knowledge.

2.2.5 Hand hygiene compliance in Zambia

Three Zambian quantitative descriptive studies on IPP have been conducted (Mukwato et al., 2008;Chiboola, 2017;Chitimwango, 2017). The two 2017 studies focused on KAP among 150 and 196 randomly sampled HCW to improve generalisability (Chiboola, 2017;Chitimwango, 2017). Descriptive statistics showed that the mean scores for knowledge, attitude and practice were 83.2, 81.4% and 48.9% respectively(Chitimwango, 2017). Inferential statistics were not used in two 2017 studies. Using a convenient sample of 77 HCWs, Mukwato et al., (2008) determined that the hand hygiene compliance of 61% was significantly associated with knowledge and positive attitude. Comparing with other LMIC studies, the compliance rate for the Zambian study was higher because it broadly looked at IPP instead of the hand hygiene subset. Furthermore, the three Zambian studies used researcher generated, validated and self-administered questionnaires instead of WHO questionnaires used in other LMIC studies hence comparison of results maybe misleading.

2.3 RESEARCH GAPS

Finally, a number of studies have been done in LMIC; the three Zambian studies were limited since they only identified the KAP towards IPP among nurses. All the Zambia studies were descriptive quantitative studies, two studies by authors Chiboola, (2017) and Chitimwango, (2017) were analytically limited because of the use of descriptive statistics without any inferential statistics thus did not check for associations among variables. The third study by Mukwato et al., (2008) went further and used inferential statistics (i.e. chi­square) to check for associations between infection prevention compliance and various predictor variables such as knowledge, attitude, management support, and infrastructure for IPP. Nevertheless, no Zambian study has identified what demographics and training factors are associated with hand hygiene knowledge among student nurses hence this research fills in the gap to solve the HAI public health problem.

CHAPTER 3 METHODS

3.1 BACKGROUND

This cross-sectional study was done to assess the knowledge and perception which student's nurses in Solwezi had towards hand hygiene. The information was required to answer the research question: ‘what demographic and training factors were associated with nursing student's knowledge of hand hygiene in Solwezi, Zambia?' Answering the research question was important to achieve the overall aim of the study.

3.2 AIM

To determine the demographic and training factors associated with nursing student's knowledge of hand hygiene in Solwezi, Zambia to ultimately provide information for action necessary to reduce HAI

3.3 OBJECTIVES

1. Review available literature about demographic and training factors associated with nursing student's hand hygiene knowledge, attitude and practices in Zambia and LMIC.
2. To collect data about knowledge and perception towards hand hygiene among nursing students using a validated WHO self-administered questionnaire.
3. To calculate knowledge scores and categorise as good, moderate or poor
4. To describe and summarise categorical data (e.g. gender) using frequency, and continuous data (e.g. age) using mean and standard deviation when normally distributed or median and interquartile range for skewed distributions
5. To use chi-square to analyse how demographics and training factors were associated with hand hygiene knowledge among Solwezi College of Nursing (SCN) students if assumptions were met and multivariate logistic analysis to determine the strength of association between various independent variables and knowledge of students
6. To use the information to inform policymakers and contribute towards initiatives aimed at improving hand hygiene techniques and make recommendations to the General Nursing Council of Zambia (GNC), and SCN based on the findings.

3.4 EPISTEMOLOGICAL APPROACH

A positivist worldview was employed for this study because it assumes an objectivist ontology where what is being measured (students hand hygiene knowledge and perception) remains relatively constant. This paradigm is known for its use of quantitative, scientific inquiry, objectivity and statistics (Khanal, 2012). The positivist approach assumes ‘parsimony', hence the hand hygiene topic was investigated using a cost­friendly study design via a quantitative (cross-sectional) survey. Use of a survey ensured feasibility of the assumption of ‘generalisable' study findings to other settings. Positivism also assumes ‘empiricism' which implied that student nurses hand hygiene data being investigated be collected using empirically verifiable means and this was achieved using a validated structured tool (Kivunja and Kuyini, 2017). Use of this tool improved objectivity and promoted better statistical analysis (Raddon, 2007). To further ensure objectivity, the researcher's positionality disregarded their feelings and other human behaviours which could have prejudiced the study (Truncellito, 2012).

Three main shortcomings of positivism include inflexibility, disregard for human behaviour, and lacks depth compared to interpretivism. An interpretivist approach would have meant a qualitative researcher integrating themselves into a society of student nurses to deeply understand the hand hygiene subject matter while routinely adjusting the data collection tool as new information arises (Tufford and Newman, 2012). Hence if the study aimed to extensively understand student nurses experiences, beliefs, feelings and meaning towards hand hygiene then using qualitative methods, an interpretivist approach would have been ideal.

3.5 STUDY DESIGN

A quantitative study design, i.e. a cross-sectional epidemiological survey using a WHO validated self-administered questionnaire was employed. The study allowed for the measurement of how multiple exposures (demographics and training factors) were associated with the (student nurses hand hygiene knowledge) outcome in Solwezi, Zambia (Gordis, 2014). As guided by the positivist approach, this study type was cost­effective and quicker to conduct in the short dissertation time frame. A notable disadvantage was generalisability because students hand hygiene knowledge and perception are unlikely to remain stagnant over time (Setia, 2016).

3.6 SETTING

The study occurred at SCN located in the urban suburbs in Solwezi city, of Solwezi district in North Western Province of Zambia (see Figure 3). The public institution was established in 1988 and has been approved by the General Nursing Council (GNC) of Zambia to train general nurses, and recently in 2018 started training public health nurses and midwives for 3 years, students graduate with a Diploma (GNC, 2017).

Figure 3 shows North Western Province (on the small map of

Zambia) and location of Solwezi district

Source: (GRZ, 2010)

Due to copyright reasons, the image was deleted by editorial staff.

SCN is under Solwezi General Hospital (SGH) administration and is located opposite SGH (another public institution) where the nursing students do their clinical attachments (GRZ, 2010). This site was chosen because of researcher proximity to the study setting and the potential to generalise results to other nursing colleges in Zambia due to similarities in student enrollment criteria, training programs offered and curriculum taught.

3.7 SAMPLING APPROACH/FRAME

The sampling frame was a total of 446 nursing student attending SCN during the 2017/2018 academic year (School Management, 2018). Using class lists, stratified random sampling via both year of study and program of study was done in Microsoft Excel 2016 with function “RANDBETWEEN(1,446)”. Stratification ensured better student representation from the two new nursing programs (i.e. public health and midwifery) which had only one intake and fewer number of students, while simple random sampling in each stratum ensured that each student had the same chances of participating in the study as other students. Overall this helped to avoid any sampling error and selection bias (Daniel and Cross, 2013). The total number of students targeted was 44, 50 and 45 from the first, second and third years respectively training as general nurses. 23 and 44 students were from the first year and only intakes for public health nursing and midwifery respectively.

3.8 SAMPLE SIZE

Based on a study by Mukwato et al., (2008) in a Zambian general hospital, given that the proportion of HCWs compliant to hand hygiene was 61%, with 5% precision the sample size is 206 nursing students. The formula used was:

Abbildung in dieser Leseprobe nicht enthalten

where Zi - a/2 is the standard normal variate, p represents the hand hygiene compliance and d is the researcher determined precision level (Bruce, et al., 2008; Charan and Biswas, 2013). Hence

Abbildung in dieser Leseprobe nicht enthalten

After accounting for a 10% non-response rate, the total sample size needed was 206 nursing students (Bruce, et al., 2008; Kadam and Bhalerao, 2010).

3.9 ETHICS

Before the study commenced, ethical clearance was obtained from Tropical Disease Research Center TRC/C4/04/2018 in Zambia and the University of Liverpool Research Ethics Committee, while gatekeeper permission was granted by SCN (see appendix). Confidentiality and anonymity were maintained by not using identifier details (names or residential addresses) and keeping records on a password protected/encrypted computer with restricted access to information only shared between the researcher and supervisor. Each participant provided written consent before inclusion in the study. Participants were notified of their rights such as being able to withdraw from the study without giving a reason and without any repercussions. All ethical research principles were guided by the Declaration of Helsinki, University of Liverpool ethics and local ethics. All research data will be stored for 5 years only.

3.10 INCLUSION AND EXCLUSION CRITERIA

The inclusion criteria for participation in the study were: nursing students of any gender at SCN pursuing any program and were 18 years and above. The exclusion criteria were: students below 18 years.

3.11 RECRUITMENT AND DATA COLLECTION METHODS

Primary data was collected for three weeks (from 17 December 2018 to 7 January 2019). Before data collection, the researcher visited the research site (i.e. SCN) to engage and sensitise management on the research objectives, the data collection tools/process, and present a copy of the local ethical approval. The researcher also confirmed the availability of the space allocated for the research. A convenient timetable of both weekly and daily activities was agreed upon with management.

Weekly Activity: The College management was notified when the researcher arrived on the premises. After seeking management's permission, students were sensitised in class about the research and provided hard copy information sheets which also contained the researcher's contact details in case of any inquiries.

[...]