Knowledge, Attitudes, Behaviour and Training Needs of Icu Nurses on Medication Errors in the Use of Iv Drugs: a Pilot Study

Background. Prior studies have shown that lack of knowledge and

Results.In total, 123 nurses completed the questionnaire.For the variable 'adequate knowledge' nurses were divided in those who had correctly answered 7 of 8 questions on knowledge and everyone else, while for 'attitudes and professional behaviour' responders were grouped into those who showed a positive attitude and appropriate behaviour and everyone else.Results of the multiple logistic regression analyses show some association, although not statistically significant, between correct behaviour, positive attitudes and the years of work experience.It is interesting to note that appropriate knowledge weighs negatively on correct behaviour (OR 0.06, 95% CI 0.04-0.42).Moreover, graduate

Introduction
Medication errors in intensive care units (ICUs) can risk patients' health or cause their death.(1) A narrative literature review estimates medication errors' percentage is included among 12% and 20% of total errors.(2) Such errors are expensive from a human, economic, as well as social point of view.(3) Under ˝medication error˝ we mean any adverse, unwelcome, preventable event which can cause or lead to an SIGNA VITAE 2016; 11 (1): inappropriate use of drugs or to a danger for the patient.Such an event can be caused by many factors, such as: a prescription error, a wrong transmission of a prescription, labelling, packing or denomination, setting up, distribution, medication, education, monitoring, as well as inappropriate use (www.nccmerp.org).( 4) Such errors have human, economic and social consequences.(5) According to many studies, medication errors in ICUs can occur as a result of patient clinical complexity, (6) the number of administered drugs, the frequent change of prescriptions, the need to fix the drug dose depending on the patient's weight, the potential incompatibility between intravenous (IV) drugs as well as the prescription speed.(7) Rothschild JM et al. claim that medication errors represent 78% of clinical errors in ICUs, and patients every day experience 1.75 errors on average during their stay in ICU.About 20% of medication errors put the patient's life in danger, while 40% require some treatment.(8) Although errors can occur at any time in the medication management process, and therefore the whole medication management system (9) must be taken into consideration, for the purpose of prevention, according to the literature, the administration phase represents the most crucial moment of the process.by the healthcare facility (organizational management aspect) and finally in terms of lost of patients' trust towards the hospital (loss of image).
According to the literature, there are four observed variables that can influence medication errors: knowledge, attitudes, behaviour and training needs.The literature states that inadequate nurses' knowledge and training on the use of IV drugs, can be reported as a cause of error.
(15) Some authors agree that wrong professional behaviour (such as incorrect handwashing) can cause medication errors; (3) the same applies to professional attitudes, which are considered by Leape as the best approach to identifying errors.(16) There are few studies that connect nurses' knowledge, attitudes,

Methods
The survey is an empirical research method in which a questionnaire (self-administered in this case) is used for data gathering and statistical methods.
Such a survey is a useful tool for acquiring information.built ad hoc on the basis of the bibliographical research, followed by a systematic literature review relevant to the research question.Questions were intended to identify demographic factors (section A), professional features (section B), continuing professional development (section C), knowledge on the use of IV drugs in ICUs (section D), attitudes on the use of IV drugs in ICUs (section E), behaviour on the use of IV drugs in ICUs (section F), as well as training needs on the use of IV drugs in ICUs (section G).As far as knowledge and behaviour sections are concerned, 8 questions were identified, while seven questions were identified in the attitudes' section.
A three-level Likert scale (agree, uncertain, disagree) was used for the knowledge and attitudes' sections (tables 2,3).Regarding the behaviour section (table 4), a four-level Likert scale (always, often, sometimes, never) was used.In the last set of questions, (G), including the section on training needs on the use of IV drugs in ICUs, the respondent nurses were asked to express their opinion about their level of knowledge on the preparation and administration of IV drugs.They had to choose among four answers (very low, intermediate, good, excellent); nurses were also asked whether, according to them, improving their knowledge with training courses would be worthwhile (the answers were "yes" or "no").
Sampling and methods of data collection: selfadministered questionnaire to ICU nurses.
The tool was subjected to validation to test the items'internal feasibility, validity and reliability.It was tested on 123 ICU nurses, working in two university hospitals in Rome.The questionnaire's reliability was evaluated by calculating Cronbach's alpha, (17) a measure of internal consistency, that is the ratio of the covariance sum of the scores of all items and the total variance of items themselves.Such a parameter can be interpreted as an average of the correlation coefficients calculated for each possible division of items into two groups of equal dimension.The assessment of a scale's reliability consists of estimating how much the score variation is real or actual, rather than being due to chance or casual errors.The reliability's degree estimated from Cronbach's alpha is expressed as a proportion: for example a 0.70 realibility degree means that the measured variance can be considered 70% reliable.( 18) The scale's length (number of items) as well as the interitem correlation (correlation between individual items) are factors that are able to modify a scale's reliability.Correlations between scales, also known as inter scale homogeneity values, above their reliability index, indicate that two scales were constructed to measure the same concept; on the other hand, inter-scale homogeneity values below their reliability index mean that separate phenomena are measured.In this way we can determine whether the scale was able to measure that for which it was built and if the results are interpretable.(19) Finally, the Bartlett's sphericity test was performed to evaluate if a principal component analysis could be performed on the dataset.

Preliminary statistical analysis
A descriptive analysis of all the questionnaire's variables was performed using absolute and percentage frequencies (tables 1, 2, 3, 4).Three models of multiple logistic regression were later constructed to identify knowledge, attitudes and behaviour predictors (table 5).For this purpose the dichotomous variable called "appropriate knowledge" was created, grouping all those who correctly answered 7 of 8 questions on knowledge into one group and everyone else in the other (Model 1).For the same purpose, the dichotomous variables "positive attitudes" and "right behaviour" were created, respectively grouping those who correctly answered all the questions about behaviour into one group and everyone else into the other (Models 2,3).The following independent variables were tested in models: age (continuous), gender, years of work, year of graduation from basic study, degree type.The "appropriate knowledge" variable was also tested as a covariate both in the attitudes' model and behaviour model.In the latter, the "positive attitudes" variable was also tested.
Multiple logistic regression models were constructed using the strategy

Ethics
The approval of the Ethics Commitee for questionnaire administration was received -Prot.44/14 PT_ComEt CBM.

Study population
The average age of the sample was 37.6 years (±9.2), with 30.1% of males and 69.9% of females.A total of 51.2% of the sample had a university degree in nursing, while the remaining 48.8% did not have a university degree.Only 16% of nurses had a first level Master's degree (table 1).

Knowledge
The majority of the sample (88.6%) considered knowledge on drug dosage calculation essential in order to reduce medication errors during the phase of drug preparation (table 2).
A total of 84.6% of the sample stated that the use of protocols, posters, as well as information brochures concerning IV drug administration on the wards, would be useful in order to reduce medication errors.
Almost all of the nurses (87%) agreed that shifts changes (double shifts, overtime, etc.) as well as workload, contribute to IV medication errors.
A relevant aspect, related to knowledge, concerns alarms and ward emergencies, which can cause distractions both in preparation and administration phases of IV drugs.For 64.2% of nurses distractions during the preparation and administration phases of IV drugs, can lead to medication errors.A total of 78.7% of the sample considered the use of a single therapy chart as a good strategy for avoiding errors during the management process of IV drugs.
According to multivariate analysis there were no statistically significant associations; the strongest associations for appropriate knowledge are related to having a university degree.

Attitudes
Nurses'answers in the attitude section are encouraging (table 3).
According to about 90% of the sample, in order to reduce errors it is necessary to take preventive measures, such as frequent trainings, authoritative guidelines which should be drawn up taking into consideration the available scientific evidence, continuous evaluation of clinical skills, as well as the error report, in order to improve care.
Almost all the nurses (90.2%) believe that awareness of prevention of errors and clinical risk management can reduce errors during the preparation and administration phases of drugs.
Multivariate analysis shows a statistically significant association between positive attitudes and feminine gender; on the contrary, appropriate knowledge is not associated with positive attitudes, even if an OR>1 was found.

Professional behaviour
For almost all the sample, basic behaviour in regard to asepsis (hand washing) have been adopted, before the preparation of therapy (table 4).
Regarding the preparation of therapy in a dedicated room, contrasting answers with the literature come to light; in fact, only 38.2% of the sample prepares IV therapy in a specific room, whereas 33% directly prepare therapy at the bedside, probably due to the absence of such rooms.
An important aspect shared by more than 90% of the sample, covers professional behaviour on the rate of infusion of IV drugs, as well as the monitoring of vital signs before and after medication administration.
Multivariate analysis results show some association, although not significant from a statistical point of view between correct behaviour, positive attitudes and increasing working experience.On the other hand it is interesting to highlight that appropriate knowledge weighs negatively on correct behavior (OR 0.06; 95% CI 0.04-0.42).

Training Needs
Nurses were asked in the questionnaire how they consider their own knowledge level about IV drug preparation and administration.Only 22.8% answered that their knowledge level is great, for 64.2% it is good, for 12.2% it is sufficient and for 0.8% it is insufficient.
Consistent with literature data (21), 95.1% consider it fundamental to improve their knowledge about IV drug preparation and administration in order to reduce medication errors.
Academic training should increase the content on the safe use of drugs.
(7) 94.3% of nurses in our study agree that good knowledge on the use of IV drugs can contribute to reduce medication errors.
One more considerable aspect is the cooperation between newly employed nurses and experienced nurses in order to prevent medication errors in ICUs.(22) Almost all respondents (99.2%) in our sample considered the presence of newly hired nurses fundamental to the prevention of errors. of nurses in our survey spend time reading the literature.
An encouraging sign concerns the percentage of nurses who find it very important to improve their knowledge on the preparation and administration of IV drugs.

Conclusions
IV administration of drugs is one of the priority responsibilities of a nurse in the ICU.It is a key moment to intercept errors that have occurred in earlier phases of drug management, before they reach the patient.
The results of this pilot study are encouraging.Experience and competence gained by nurses in ICUs is crucial for implementing appropriate behaviour in order to prevent medication errors.(35) One more factor emerging from this study is the need for a knowledge-based clinical practice, as well as theoretical knowledge on IV drugs.
Experience plays a key role, probably because it creates awareness of the need for respecting correct behaviour throughout the process in order to prevent medication errors.
Graduate nurses seem to have general theoretical knowledge, but not atdeeper levels to be able to use correct behaviour in specialized areas such as ICUs.(36) The questionnaire will be amended, especially in the knowledge and SIGNA VITAE 2016; 11(1): behaviour sections, where the Cronbach's alpha was less than 0.7.It will probably be related to the type of questions as well as the response method.
The number of involved nurses in the study is not high.Such contribution is only the starting point of an Italian multi-center study which aims to stress the predictors of medication errors in ICUs.

( 10 )
The act of medication administration represents 40% of nurses'clinical care activities.(11) Errors caused from multiple administrations are particulary common in ICUs.(12) Critical patients receive almost twice as many drugs in comparison to hospital patients staying in other operating units.(13,14) Most drugs involve some calculation of bolus administration or continuous infusion administration.Nowadays, reliable data related to medication errors in adults hospitalised in ICUs, are not yet available.Patient safety is an issue in all advanced health systems.Such concern is highlighted and shared by the World health organization (WHO) as well, which, since 2001, asks for urgent action by health organizations themselves (Ministry of Health, 2012).(5) Adverse events are far more numerous than those supposed or than those which actually damage the patient.Such events inevitably lead to poor quality care, in terms of effectiveness (clinical appearance), in terms of increased costs incurred SIGNA VITAE 2016; 11(1): behaviour and training needs to the administration errors of IV drugs.Based on our literature review, there are no tools to assess whether nurses' knowledge, attitudes, behaviour and training needs about the use of IV drugs in ICUs determine any medication errors.The aim of this study was to describe ICU nurses' knowledge, attitudes, behaviour and training needs on the use of IV drugs in ICUs to prevent errors.
The questionnaire is preceded by a brief description in which details and the aim of the study are explained, protecting anonimity of participants (informed consent).It is a cross-sectional study.The questionnaire consisted of 43 items as well as seven sections.It was SIGNA VITAE 2016; 11(1): SIGNA VITAE 2016; 11(1): suggested by Hosmer and Lemeshow (20) including the following phases: (a) univariate analysis of each variable; (b) Odds ratio (OR) and its related 95% Confidence Interval (CI).For statistical analysis the SPSS ® software (Statistical Package for Social Science) for Windows, version 22.0 was used.

Table 1 .
Demographics and professional and personal characteristics of the responding nurses.

Topics related to the preparation and administration of IV drugs covered during the postbasic course (84) a
a Number of nurses responding to the question.

Table 2 .
Knowledge on the use of IV drugs of the responding nurses in a Number of nurses responding to the question.IV, intravenous.SIGNA VITAE 2016; 11(1):

Table 3 .
Attitudes on the use of intravenous (IV) drugs of the responding nurses in Intensive care units (ICUs).
a Number of nurses responding to the question.

Table 4 .
Behaviour on the use of IV drugs of the responding nurses inIntensive care units (ICUs).
a Number of nurses responding to the question.

Table 5 .
(27)ictors of nurses' knowledge, attitudes and behaviour on the prevention of medication errors.Abstoss KM, et al.(27)stated that the most implemented and recommended action to collect and record such alerts is the application of an incident reporting system.This is also confirmed by our study, which showed that 91.8% of nurses consider fundamental to report medication errors, in order to improve health care.VITAE 2016; 11(1): is associated with high competence and has a greater preventive effect on medication errors.(34)Multivariate analysis' results confirm that graduate nurses have more appropriate knowledge than those who are not.As a matter of fact, such appropriate knowledge is not related to appropriate behaviour and positive attitudes on management of IV drugs.Knowledge about prevention of medication errors could be improved with specific training courses and with an increase in the time (hours) spent on reading the scientific literature.Only a small percentage (35%) Multivariate analysis stresses an interesting phenomenon: correct behaviour is associated with inadequate knowledge.This is probably due to the fact that, in a specialized area such as ICUs, long work experience SIGNA