Relationships between the Advanced Nursing Process quality and nurses' and patient' characteristics: A cross‐sectional study

Abstract Aim This study aimed to assess (a) nurses' knowledge and their attitude towards the Advanced Nursing Process—nursing assessment, diagnoses, interventions, outcomes, (b) the quality of the Advanced Nursing Process and (c) relationships with patient characteristics. Design A cross‐sectional, descriptive correlational study was performed. Methods Ninety‐two registered nurses and ninety nursing records of six hospital wards were included. In January 2016, a knowledge test, a self‐assessment tool for measuring nurses' attitude (PND) and the Quality of Diagnoses, Interventions and Outcomes Revised instrument (Q‐DIO R) were applied. The correlations between nurses' knowledge, attitude, patient characteristics, organizational factors and the Advanced Nursing Process quality were investigated. Results Nurses demonstrated low levels of knowledge, positive attitudes and an average Advanced Nursing Process quality. Accurate nursing diagnoses were strong and highly significantly related to effective nursing interventions and better nursing‐sensitive patient outcomes. A higher proportion of registered nurses was related to better nursing outcomes.


| Background
The Advanced Nursing Process is a specific form of the "traditional nursing process." It consists of valid assessment tools and on defined, valid concepts of nursing diagnoses (NDs), nursing interventions and nursing outcomes "that are rooted in scientifically based nursing classifications" (Müller-Staub, Abt, Brenner, & Hofer, 2015, p. 13).
The NNN taxonomy (Johnson et al., 2011), consisting of the nursing diagnoses classification of NANDA International, the Nursing Outcomes Classification (NOC) and the Nursing Interventions Classification (NIC), meets most of the validity and reliability criteria for classifications systems (Müller-Staub, Schalek, & König, 2017).
The quality of the Advanced Nursing Process is not only determined by its documentation in nursing records, but rather by its internal coherence. Coherence means that accurately formulated NDs are correctly linked with effective nursing interventions and matching outcomes. The accuracy of NDs relies on the PES format (problem definition, aetiological factors and signs/symptoms) and is the initial point for choosing effective nursing interventions and outcomes showing the effectiveness of interventions. The quality of NDs also depends on their relevance, frequency and variety to describe patients' care needs correctly. Fewer NDs and a smaller variety could depend on the patient and the reason for hospital admission or/and could also indicate lower diagnostic quality (Johnson et al., 2011;Müller-Staub et al., 2009).
The quality of the Advanced Nursing Process is influenced by different factors: firstly, by nurses' knowledge and attitude. For the implementation of the Advanced Nursing Process, nurses need patient-related knowledge about the person and situation, clinical decision-making competence including diagnostic reasoning and knowledge about validated concepts of the NNN taxonomy (Lunney, 2007;Paans, Nieweg, Schans, & Sermeus, 2011). The professionals' behaviour (the implementation of the Advanced Nursing Process) is influenced by their individual intention to act. This intention, in turn, is mainly determined by nurses' attitude referred to their behaviour (Ajzen, 2012). Secondly, the Advanced Nursing Process is influenced by patient characteristics such as age, admission to medical speciality and LOS. Further, also organizational factors could influence the Advanced Nursing Process quality. Several studies addressed the relationships between nurses' knowledge and attitude towards the Advanced Nursing Process (Kim & Shin, 2016;Paans et al., 2011), respectively, nurses' knowledge or attitude and organizational factors (Lumillo-Gutierrez et al., 2018;Okaisu, Kalikwani, Wanyana, & Coetzee, 2014) and connections between knowledge, attitude and organizational factors (Kebede, Endris, & Zegeye, 2017). Studies focusing on all these relationships with the entire Advanced Nursing Process-nursing assessment, diagnoses, interventions and outcomes-are missing.

| Aims and research questions
The overall objective was to measure the effects of an educational intervention on nurses' knowledge, attitude and the Advanced Nursing Process quality by performing an experimental intervention study (Leoni-Scheiber, Mayer, & Müller-Staub, 2019). The aim of this paper is to report baseline findings and relationships between nurses' knowledge and attitude towards the Advanced Nursing Process; the quality of NDs, interventions and outcomes; and patient characteristics (gender, age, LOS).

| DE S I G N
A cross-sectional, descriptive correlational design was carried out to investigate the relations between nurses' and patient characteristics and the quality of NDs, interventions and outcomes.
Therefore, a knowledge test for nurses, a self-assessment instrument measuring nurses' attitude towards the Advanced Nursing Process and a tool for demographic and organizational factors were applied. The Advanced Nursing Process quality was measured by using the Q-DIO R instrument. The study was conducted in a Swiss 260 bed hospital in three departments (medical, surgical, acute geriatric).

| Sample/Participants
Nurses and nursing records of six hospital wards were included. The inclusion criterion was general wards (two medical, two surgical and two acute geriatric). In January 2016, using stratified convenience sampling, a third of the RNs from these wards (N = 34) was selected and asked to fill out the knowledge test. Eligibility criteria for nurses were as follows: holding a position as ward manager; nurse instructor; Advanced Nursing Process mentor; or being a regular RN applying the Advanced Nursing Process covering all shifts and speaking German fluently. The composition of this group based on her specific responsibility for the Advanced Nursing Process. Later, this group would participate in an educational intervention study. All RNs of the six wards (N = 99), not only the group described above, were included for measuring their attitude towards the Advanced Nursing Process. The reason for including all nurses in the attitude measurement was that all RNs would later participate in case meetings on their wards for a following study.
The quality of the Advanced Nursing Process was evaluated by nursing record audits of the six wards in January 2016 (15 per ward, in total 90 records). The nursing process documentation was part of the Electronic Health Record (EHR) that included 43 NANDA-I NDs with related nursing outcomes and interventions formulated according to Doenges, Moorhouse, and Murr (2014). A decision support system was not available in the EHR at that time. The nursing records of each patient were created by the responsible RN at any shift. Therefore, it is a cumulative product created by several nurses.
Besides the nursing assessment and care plan, other nursing record content (e. g. nursing notes/report, performance descriptions)-of the duration of four days after stating the Nursing Diagnosis-was included. Record inclusion criteria were as follows: containing an individual care plan with at least one ND and patients' LOS for at least 4 days. A random sample of records was drawn by random generator (https ://www.random.org/).

| Measurement instruments
A knowledge test was developed and pilot-tested in a previous educational training. After minor adjustments, it contained seven items: Six qualitative knowledge questions (e.g. definition of nursing diagnosis, meaning and function of the PES format, coherence between NDs, interventions and outcomes) and one self-evaluation item on nurses' knowledge. Results of this self-evaluation item were compared with the total score of the knowledge test (maximum 58 points).
Nurses' attitude towards NDs was measured by the Positions on the Nursing Diagnoses (PND) scale (Lunney & Krenz, 1994). It contains 20 contrasting pairs of attitude adjectives on a seven-point Likert scale (e.g. easy-difficult, important-unimportant). Its total score ranges from 20-140; the more positive the attitude, the higher the total score. The German PND (Lunney, 2007) was previously evaluated in a quasi-experimental study (Leoni-Scheiber, Gothe, & Müller-Staub, 2016). The original PND was repeatedly tested and showed good results, for example content validity 90 per cent agreement (Lunney & Krenz, 1994), test-retest reliability 0. To assess the quality of the Advanced Nursing Process, the Quality of Diagnoses, Interventions and Outcomes (Q-DIO) instrument (Müller-Staub et al., 2009, 2010 was revised and applied for the record audit. The Q-DIO R was expanded from 29-35 items, for example the item "congruence between assessment and nursing history data and the complexity of the case" has been added. The scales remained, a three-point scale (0-2), was used for the twelve nursing assessment items, a five-point scale (0-4) for the other three subconcepts (NDs as product, nursing interventions and nursing outcomes). The higher the total score of each sub-concept, the higher their quality. The original Q-DIO scale was validated in several studies (Linch et al., 2015;Müller-Staub et al., 2009, 2010. Internal consistency by Cronbach's alpha ranged from 0.83-0.99 for each sub-concept, test-retest reliability by kappa was 0.95 and interrater reliability by Fleiss' kappa was 0.94. Standardized data collection charts were used to determine potential confounders. Hereby, data on nurses' gender, their highest education grade, years of practical experience and previous Advanced Nursing Process education were collected. From each ward, data on bed occupancy rate, patients' LOS, nurse-to-patient ratio (calculation based on full-time equivalents), skill-and grademix, staff turnover and characteristics of the nursing care delivery system (e. g. autonomy and authority about nursing decisions, communication) were collected.

| Procedure
Nurses completed the PND and the knowledge test on their wards and ward managers/Advanced Practice Nurses gave instructions on applying the test (e. g. answering all questions, time limit, submitting). After explaining, first the PND was silently completed by hand, requiring five minutes. The PND's were coded using a personal fourdigit number (e.g. last phone digits) to assure anonymity and comparison with their knowledge tests. Second, the knowledge test was silently taken by handwriting. Its completion allowed a maximum of 30 min. The participants were asked to fill in the same four-digit number as before. The response sheets were collected in a large envelope and closed by the last participant. The knowledge tests were assessed by the first author according to standardized responses and standard scores as outlined by the author of the test.
In the nursing records, each of the 35 Q-DIO R items was evaluated by the first author. To ensure a consistent approach, the instrument developer also evaluated several records and comparisons were made. If evaluations differed, consensus was sought and correct instrument application was assured by writing memos.
Nurses' characteristics were collected with the PND, patient characteristics from the nursing records and the principal investigator charted organizational data from all six wards.

| Data analysis
All data were analysed by IBM SPSS Statistics 24 (SPSS Inc.).
Descriptive statistics were performed for frequencies and distribution of sample characteristics and outcome measures (nurses' knowledge, attitude, quality of the Advanced Nursing Process). Inductive statistics (Kruskal-Wallis tests) were used to compare the outcome measures between the three departments (medical, surgical and acute geriatric). Repeated post hoc tests between groups were conducted including Bonferroni correction (Clauß, Finze, & Partzsch, 2011). Comparisons based on single items and on total scores of the knowledge tests and self-assessments of nurses' attitude (PND).
The Q-DIO R scores were summed up for each sub-concept (Müller-Staub et al., 2010). Pearson's product-moment correlations were used to determine the relations between metrical data (knowledge scores, Q-DIO R scores, patient characteristics, e.g. age and LOS and organizational factors). The relations between ordinal data (nurses' attitude, educational degree and practical experience) or between ordinal and metric data were analysed by Spearman correlations.
Chi-squared tests were used to investigate associations between patients' gender and for previous Advanced Nursing Process education with other variables (LOS, knowledge, attitude and Q-DIO R subconcepts). Linear regression analyses were performed to determine how well a dependent variable (e. g. quality of nursing outcomes) could be explained by independent variables (e. g. accuracy of NDs, effectiveness of nursing interventions) if the statistical prerequisites were attained. All tests were applied one-sided (significance level 0.05) (Bortz & Schuster, 2016).

| Ethical considerations
Ethics approval was obtained from the responsible cantonal committee (Nr. PB_2016_00990). The study was executed in conformity with the Declaration of Helsinki (WMA, 2013) and with the study country's laws and regulations. The nurses from the six hospital wards were asked for voluntary participation and gave informed consent.

| Validity and reliability/Rigour
As described in the instrument section, the PND and Q-DIO were psychometrically tested and showed good results. In a pilot application, the knowledge test was evaluated by experts and modified before application. Additionally, the research team was trained to assure standardized data collection and the STROBE Statement (Appendix S1) was used for reporting (Elm et al., 2007).

| RE SULTS
Ninety-nine RNs were included, 92 returned the PND (response rate: 92.9%). Most of the nurses were female, had a tertiary (academic) degree and one third had more than 21 years of practical experience (Table 1). Patient characteristics in nursing records and departments are shown in Table 2. Two thirds of patients were female, acute geriatric patients were on average older, and their average LOS was nearly doubled compared with medical and surgical patients.  Research question 2: What is the quality of the Advanced Nursing Process as measured by (a)  Expected patient outcomes were achieved and demonstrated intervention effectiveness with 1.9 (SD 0.6) (range: 0.5-3.6). The highest rated quality of NDs, interventions and nursing-sensitive patient outcomes were attained on both medical wards (Table 3). There were significant differences in the quality of nursing interventions between the medical and surgical wards (p = .014) as well as in the quality of nursing outcomes between the medical and acute geriatric wards (p = .002).
The internal coherence and thus the correlation between NDs, interventions and nursing-sensitive patient outcomes was strong and highly significant. The higher the accuracy of NDs the more effective The standardized regression coefficients (ß) show that the variance of the quality of nursing outcomes can be explained by the accuracy of NDs more than by the effectiveness of nursing interventions. In  Further, we found out that the skill-and grade-mix was statisti- of the other organizational factors (Table 5) were statistically significantly related with the ND' quality.

| Synopsis of main results
The overall analyses of knowledge and attitude tests, as well as of nursing records demonstrated rather low levels of nurses' knowledge of, a positive attitude towards and an average quality of the Advanced Nursing Process. The quality of the nursing assessment and ND was slightly above the Q-DIO score average, while the quality of nursing interventions and outcomes were slightly below the mean. A strong relationship was found between the quality of ND, interventions and outcomes. The more accurate the ND, the more effective the interventions and the better the nursing-sensitive patient outcomes. This demonstrates that accurate NDs are key and the starting point for the next phases of the Advanced Nursing Process (e.g. choosing effective interventions to attain favourable patient outcomes). The nurses of the medical wards had more knowledge, a slightly more positive attitude and the quality of ND, interventions and outcomes as well as the variety of NDs were the highest compared with the other wards. The higher the proportion of RNs within a team was, the better were nursing-sensitive patient outcomes.

| Nurses' knowledge and attitude
When looking specifically at these variables, nurses' knowledge towards the Advanced Nursing Process was limited. A partial explanation may be that almost half of the RNs' basic education dated back more than 15 years. RNs of Greek primary healthcare settings with similar years of experience (Patiraki et al., 2017) and Ethiopian nurses (Kebede et al., 2017) also showed lack of knowledge. In contrast, nursing Bachelor students (more than half of them had a previous nursing diploma) showed good knowledge about using NDs

| Quality of ND, interventions and outcomes
The mean Q-DIO scores of the quality of NDs (x = 2.3), interventions (x = 1.9) and outcomes (x = 1.9) ranked in the lower midfield Nursing Process quality was found on medical wards compared with surgical ones. In contrast, in a Nigerian university hospital the Q-DIO total scores were significantly higher in surgical than in medical wards (Adubi, Olaogun, & Adejumo, 2017).
We identified a strong and highly significant correlation between the quality of ND, effectiveness of interventions and better nursingsensitive patient outcomes. Other studies support these results. In a Brazilian hospital, where NANDA-I and NIC were used, the accuracy of ND, the effectiveness of nursing interventions and the outcomes were significantly better compared with another hospital where the International Classification for Nursing Practice (ICNP) was applied (Rabelo-Silva et al., 2016). The ICNP does not contain a comprehensive description of concepts (PES format) and theory-based linkages between ND and nursing interventions and outcomes are missing (Müller-Staub et al., 2017). The quality of nursing outcomes was also influenced by the number of NDs: fewer NDs were statistically significantly associated with better nursing-sensitive outcomes (Scherb et al., 2013) while patients with a higher number of NDs tended to be in worse general conditions (Castellan, Sluga, Spina, & Sanson, 2016;D'Agostino et al., 2017). number of NDs per patient in our study was 2.6. All previous studies showed more than 3.5 [from 3.8 (Aguilar & Pancorbo-Hidalgo, 2011) to 9.6 (Araúja, Carvalho, & Chianca, 2014]. The variety of NDs depended on the number of NDs that were contained in the EHR.

| Frequency and variety of NDs
In the EHR that was used in the study site, only 43 NDs-out of totally 245 NANDA-I NDs-were available. From these, 31 were used, twelve were not used and only two additional ones were stated. In a study with elderly patients undergoing surgery, 14 of 74 possible NDs were used (Tuncbilek & Celik, 2016

| Correlations between nurses' knowledge, attitude and the quality of the Advanced Nursing Process
Nurses' knowledge was associated with the quality of the Advanced Nursing Process. After training, nurses identified more accurate NDs (Pobocik, 2015;Predebon et al., 2012). Other factors, such as a predefined record structure with the PES format, were significantly associated with more accurate NDs (Paans, Sermeus, Nieweg, Krijnen, & Schans, 2012); and three studies reported that participants of nursing process education had a more positive attitude towards NDs (Collins, 2013;D'Agostino et al., 2016;Romero-Sanchez et al., 2013).
Higher knowledge on and positive attitudes towards the Advanced Nursing Process were associated with better nursing process quality (Kebede et al., 2017). In our study, solely the nursing assessment quality revealed a positive, significant correlation with nurses' attitude.

| Correlations between patient characteristics and the quality of the Advanced Nursing Process
Higher patient age and longer LOS were associated with a higher number of NDs per patient and less positive nursing-sensitive outcomes. In previous studies, nurses stated also more NDs in elderly patients [6.1 (Heering, 2011)

| Limitations
The results of this study are transferable to a limited extent because of the small sample size (knowledge test) and the sampling procedure (convenience sampling for knowledge tests). The reasons for sampling were the size of the hospital and the availability of organizational resources for this study. Due to missing statistical prerequirements (e. g. normal distribution of the dependent variable, multicollinearity), multilevel regression analyses could be applied not for all variables. Despite this, the sample size of nursing records, the multiple foci of measurements and the comparability of departments allow reliable conclusions.

| CON CLUS IONS
The aim of this study was to detect possible relationships between nurses' knowledge, attitude, patient characteristics and the Advanced Nursing Process quality. Nurses' knowledge and positive attitudes are key for a good Advanced Nursing Process quality. Better nursing assessments, more accurate NDs and better nursing-sensitive patient outcomes were associated with good knowledge. Nurses need good knowledge to state NDs in the correct PES format, to derive appropriate nursing interventions from the NDs' aetiological or risk factors and to link them with relevant and fitting patient outcomes. The measured quality of nursing assessment, diagnoses, interventions and outcomes was on average, compared with other studies. We conclude that nurses should be supported to develop their knowledge and clinical decision-making competencies.
Relationships between patients' higher age, longer LOS, higher number of NDs and less favourable nursing-sensitive patient outcomes provide clues on the severity of patients' health conditions.
In hospitals, there is a tendency for increasing numbers of elderly patients in complex situations. Furthermore, the quality of nursing assessments and patient outcomes was better in wards with higher proportions of RNs. These factors should be considered, as it is known that good nurse-to-patient ratios are leading to better nursing care.
To our knowledge, this is the first correlational study reporting strong relationships between the quality of NDs, intervention effectiveness and better nursing-sensitive patient outcomes. Further research using diverse research methods, larger samples in different settings and additional statistical evaluations such as multilevel regression analyses are warranted to address relations and effects when applying the Advanced Nursing Process.

ACK N OWLED G EM ENT
We express our deepest gratitude to the nursing director of the study site Ms Alexandra Heilbronner and to our statistician Ms Martina Koller.

CO N FLI C T O F I NTE R E S T
No conflict of interest has been declared by the author(s).