Accepted: 19 October 2017 DOI: 10.1111/jocn.14135 ORIGINAL ARTICLE The impact of primary nursing care pattern: Results from a before–after study Alberto Dal Molin PhD, MSN, RN, Researcher, Coordinator1,2 Head Nurse of Medical Units 3 | Claudia Gatta MSN, RN, | Chiara Boggio Gilot MSN, RN, Project Coordinator4 | Rachele Ferrua MSN, RN, Nurse of Cuneo Hospital and Advisor4 | Tiziana Cena MS, Biostatistician5 | Marie Manthey MSN, RN, Founder and Emeritus President6 | Antonella Croso MSN, RN, Director of Nursing7 1 Department of Translational Medicine, University of Piemonte Orientale, Novara, Italy 2 Nursing School, Biella Hospital, Biella, Italy 3 Biella Hospital, Biella, Italy 4 Aims and objectives: To explore the effect of “Primary Nursing” on nursing-sensitive patient outcomes, staff-related outcomes and organisation-related outcomes. Background: Primary nursing is one example of a care pattern that has recently been implemented in many countries. Centro Studi Professioni Sanitarie (CESPI), Torino, Italy Design: Before–after study. 5 Unit of Medical Statistics and Cancer Epidemiology, Center of Oncological Prevention (CPO), University of Piemonte Orientale, Novara, Italy Methods: This study was conducted in an Italian hospital. We observed patient- 6 3,169 inpatients after its implementation. Staff-related outcomes (nursing compe- Creative Health Care Management, Minneapolis, MN, USA 7 Nursing and Midwifery Areas, Biella Hospital, Biella, Italy Correspondence Alberto Dal Molin, Universit
a del Piemonte Orientale, Novara, Italy. Email: alberto.dalmolin@med.uniupo.it related outcomes (pressure ulcers, falls, urinary tract infection and venous catheter infection) in 2,857 inpatients before the implementation of primary nursing and in tence and diagnostic thinking) and organisation-related outcomes (empowerment and team climate) were collected from 369 nurses. From a subgroup of inpatients, we collected data regarding their satisfaction with the care provided. Results: After the implementation of primary nursing, some nursing competencies and diagnostic thinking were improved, so were organisation-related outcomes. Our data showed that the number of inpatients with urinary catheter infections (5.5% Funding information This research project was supported by a grant from the “Fondazione Cassa di Risparmio di Biella.” vs. 4.3%) and venous catheter infections (peripheral: 2.2% vs. 1%; central: 5.6% vs. 1%) was significantly decreased; also, the numbers of falls (2.4% vs. 1.9%) and pressure ulcers (4.8% vs. 4%) decreased, although these decreases were not statistically significant. Overall, the implementation of primary nursing increased patient satisfaction with nursing care (193.57 vs. 210.21). Conclusions: Primary nursing improved staff-related outcomes, nursing-sensitive patient outcomes and organisation outcomes. Relevance to clinical practice: Our results show that primary nursing has the potential to positively impact on inpatients, nurses and organisations. Moreover, the implementation of this care pattern is feasible. KEYWORDS nursing care pattern, nursing-sensitive patient outcomes, primary nursing, satisfaction with care, staff-related outcomes 1094 | © 2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jocn J Clin Nurs. 2018;27:1094–1102. DAL MOLIN | ET AL. 1095 1 | INTRODUCTION A large number of nursing care patterns have been introduced in many countries in recent years. Nurse staffing is associated with patient care quality and patient outcomes (Butler et al., 2011). A What does this paper contribute to the wider global clinical community? • patterns like primary nursing in daily practice because well-defined model of care is important because it enables nurses in our results show that it has the potential to positively an organisation to envision and communicate their practices (Slatyer, Coventry, Twigg, & Davis, 2015). We can define a model of care as a set of structures and processes that support nurses’ control over their practice (Hoffart & Woods, 1996). Our study suggests the implementation of nursing care impact on quality of care and patient satisfaction. • When a new care pattern is implemented, patient outcomes must be analysed. Although team/functional nursing, primary nursing (PN) and patient-focused care are the main care delivery models examined in the international literature, other models or care patterns that focus care were not different in hospitals where different models or on professionals have also been developed. However, few studies care patterns were implemented (team nursing, case manage- have explored the efficiency of such models and their impact on ment and PN). Melchior et al. (1996) further showed that nursing staff, patients and organisations (Barelli, Pallaoro, Perli, Strim- although the level of burnout did not change among psychiatric mer, & Zattoni, 2006). nurses after PN was applied, job turnover decreased. Other data As a nursing care pattern, PN aims to improve the quality of showed that primary nurses experienced more autonomy and care and nursing professionalism (Barelli et al., 2006). More specif- worked harder under this care model (Melchior et al., 1999). In ically, one nurse, termed a “primary nurse,” is responsible for the another overall episode of care for some patients (Butler et al., 2011; were reported after the implementation of PN (Boumans & Manthey, 2002). The PN therefore holds the nursing care respon- Landeweerd, 1999). study, no significant changes in absenteeism sibility for assigned patients throughout the length of their hospi- In a recent Cochrane review, in which PN was compared with talisation. The primary nurse assesses inpatients, coordinates nurse the usual model of nursing (such as functional nursing), it was indi- diagnosis, schedules care by defining priorities and, if necessary, cated that nursing staff turnover might improve as a result of PN develops an individualised discharge plan (Manthey, 2002). When application (Butler et al., 2011). However, the authors concluded the primary nurse is off duty, other nurses (termed “associate that because few studies have been conducted on PN, the evidence nurses”) provide care in accordance with the care plan. This pat- remains very limited. tern ensures continuity of care and can be applied to all types of Up to now, no studies have yet analysed the impact of PN on medical services (Barelli et al., 2006). In contrast, in other models patient outcomes; because changes in nursing staff have important or patterns (such as team/functional nursing), nurses may concen- implications on healthcare provision, it is important to plan a large trate on their tasks instead of focusing on inpatients’ needs study in which staff, organisation and patient outcomes are consid- (Barelli et al., 2006). ered (Butler et al., 2011). Primary nursing can be described as a delivery system comprising To evaluate the impact of a new care pattern, some nursing-sen- four organisational elements which differentiate PN from other sys- sitive patient outcomes must be analysed, for example, infections, tems, such as functional nursing, team nursing. The four elements falls, pressure ulcers, complications or medical errors (Butler et al., are as follows: (i) responsibility for relationships and decision-making; 2011). In this study, nursing-sensitive patient outcomes were defined (ii) work allocation and patient assignments; (iii) communication as “those that are relevant, based on nurses’ scope and domain of among staff members; (iv) management and leadership philosophy practice, and for which there is empirical evidence linking nursing (Wessel & Manthey, 2015). inputs and interventions to the outcomes” (Doran et al., 2006). An assessment of nurses’ competence—their diagnostic ability in particular—is also crucially important. A nurse’s main concern is helping 2 | BACKGROUND people, families and communities to be as healthy as possible. Before a nursing intervention is to be applied, nurses are expected The effect of the implementation of PN has been analysed in to develop a nursing diagnosis, that is, to identify the problem or risk previous studies. In an observational study comparing PN with state (Lunney, 2010). team nursing, the authors indicated that the cost per patient per Moreover, some organisation outcomes have to be investigated day decreased in wards where PN has been implemented (Gard- when a new care pattern is introduced. Team climate is relevant fac- ner & Tilbury, 1991). According to another study, inpatients tor to study because a team that works well is a requisite to provid- who could identify a nurse managing their care reported more ing adequate patient care (Wensing, Wollersheim, & Grol, 2006). positive experiences (Thomas, McColl, Priest, & Bond, 1996). Furthermore, the empowerment of both team and leader is worth However, Kangas, Kee, and McKee-Waddle (1999) noted that investigating because an empowered team can improve the flexibility nurses’ job satisfaction and patients’ satisfaction with nursing and efficiency of an entire organisation (Arnold, Arad, Rhoades, & 1096 | Drasgow, 2000). Another parameter that has to do with patients’ outcomes is length of hospitalisation whose data provide a valid proxy for estimating the consumption of hospital resources (Papi, Pontecorvi, & Setola, 2016). DAL MOLIN ET AL. • Inclusion: Nurses working in the wards where this care pattern • Exclusion: Nurses absent for over 2 months as well as those who was implemented. had not attended specific training sessions. Setting up a new nursing care pattern is a complex intervention because many organisational levels are targeted by the intervention The study began in March 2013 and ended in August 2014. (the head nurse, nurse and patients), and a large number of outcomes (patient, organisation and staff outcomes) and a wide range of interacting components get involved (Craig et al., 2008). Addition- 3.3 | Procedures ally, some interventions (e.g., education) are necessary to support The model was implemented involving nurses in the organisational the introduction of the care pattern. The Medical Research Council change. has suggested a framework that can help researchers to structure To aid all nurses involved to familiarise themselves with PN, an adequate methods for evaluating the efficiency of complex interven- online course and a 4-day workshop were set up. The syllabi cov- tions (Campbell et al., 2000; Craig et al., 2008, 2013). In the imple- ered the origins, principles and application of PN. mentation of a new care pattern, the roles of head nurses and nurse During the implementation, focus groups were created to iden- managers are primarily important because it is vital that the tify major strengths and weaknesses. A group of expert nurses with managers adapt to the changing environment (Drach-Zahavy & adequate knowledge in PN supported the implementation of the Dagan, 2002). project within the hospital. We ensured that (i) every patient was At our study site, nursing practice was aligned with Virginia Hen- allocated a nurse who was responsible for their nursing care (primary derson’s need theory (Matt, 2014) and this theory provides a base nurse); (ii) an individual nursing care plan had been drawn up; and for patient-centeredness which also is the ultimate goal of PN. (iii) the discharge plan was set, if necessary. The aim of this study was to explore the effect of PN on patient-, staff- and organisation-related outcomes. We considered the implementation of the model in a unit to be completed when at least 80% of the patients admitted were taken in charge by a primary nurse within 24–48 hr after accessing the 3 | METHODS service. 3.1 | Research setting 3.4 | Outcomes—data collection tool The study was conducted in a 400-bed community hospital in Pied- 3.4.1 | Patient-related outcomes mont, Italy. In this setting, PN was implemented in all units except for the Accident and Emergency (A&E) Department, the Outpatient The following nursing-sensitive patient outcomes (Butler et al., Clinic, the Sterilisation Station, and the Operating Theatre. Previ- 2011) were measured: ously, a functional care pattern had been used. • 3.2 | Study design and sample before (t0) and after (t1) the implementation of PN. To explore the lected data from patients and nurses. Advisory Panel and National Pressure Ulcer Advisory, 2009). We • • Eligibility criteria for patients (patient-related outcomes): • • Patients in units where PN was implemented; ≥18 years old. sue, usually over a bony prominence, as the result of pressure or pressure in combination with shear (European Pressure Ulcer This was a before–after study in which the variables were measured effect on patient-, staff- and organisation-related outcomes, we col- Pressure ulcer: localised injury in the skin and/or underlying tis- • considered both onset and staging; Patient fall: falls irrespective of cause; Urinary tract infection: inflammatory responses of the epithelium of the urinary tract to microbial invasions. Onset was determined according to clinical signs and urine culture; Venous catheter-related infection: infections resulting from the use of venous catheters. Onset was determined according to clinical signs and blood culture. Eligibility criteria for patients (satisfaction with care): • • • To gather this data, we have introduced a data collection tool in Patients in units where PN was implemented; the ward before and after the implementation of PN. These out- ≥18 years old; comes were collected by staff nurses adequately trained during hos- Able to understand and provide consent for participation. pitalisation and discharge. Eligibility criteria for nurses (staff-, organisation-related out- and after (from December 2013–May 2014) the implementation of These data were collected before (from May–November 2013) comes): PN. DAL MOLIN | ET AL. 1097 Rattazzi, & Muraro, 2007). This tool was used to analyse head 3.4.2 | Satisfaction with care nurse leadership and, in particular, the following five factors: Another outcome measured was satisfaction with care (the extent to learning by example, participative decision-making, coaching, which patients valued the health care service, in particular nursing informing and showing concern/interacting with the team. The care). These data were collected: • items were self-rated by the nurse. Team climate was measured using the Team Climate Inventory • Before the implementation of PN: second week of April, June • and August 2013; Baiardi, Zotti, Anderson, & West, 2002), which takes fives factors. After the implementation of PN: second week of April, June and The questionnaire included 38 items, and the exploratory and August 2014. confirmatory factor analysis indicated that there were five (TCI) (Anderson & West, 1998; Ouwens et al., 2008; Ragazzoni, factors. Cronbach’s alpha ranged from 0.83–0.93 (Ouwens et al., All patients meeting the eligibility criteria completed the New- 2008). The results of the Italian validation showed that a five- castle Satisfaction with Nursing Scales (Piredda et al., 2007; Thomas, factor McColl, Priest, Bond, & Boys, 1996). The questionnaire was com- Cronbach’s alpha was lower than that assessed in the original pleted by the patients without the aid of the clinical staff. The Italian investigation (Ragazzoni et al., 2002). structure was appropriate for the Italian sample. Newcastle Satisfaction with Nursing Scale exhibited very good reliability (Cronbach’s a was 0.94); face validity was also analysed (Piredda Staff and organisation outcomes were collected by the teachers/ et al., 2007). tutors during the training course. 3.4.3 | Staff-, organisation-related outcomes 3.5 | Statistical analyses The following staff-related outcomes (Butler et al., 2011) were mea- A descriptive statistical analysis was performed; means and standard sured: deviations as well as absolute and relative frequencies were calcu- • lated. Nursing competence: This outcome was analysed using the Nurs- The data were compared using Student’s t tests (for quantitative ing Competence Scale (Dellai, Mortari, & Meretoja, 2009; Finotto variables) and chi-squared tests (for qualitative variables). For all t & Cantarelli, 2009; Meretoja, Isoaho, & Leino-Kilpi, 2004), which test, Cohen’s D effect size was calculated with online calculator comprises 73 items, each of which is self-rated by the nurse. (https://www.psychometrica.de/effect_size.html#cohen): if Cohen’s There are seven groups of competencies: “helping role,” “teach- D was between 0–0.19, we considered no effect; between 0.2–0.49 ing-coaching,” “diagnostic functions,” “managing situations,” “ther- small effect; between 0.5–0.79 intermediate effect and more than apeutic interventions,” “ensuring quality” and “word role.” The 0.80 large effect (https://www.psychometrica.de/effect_size.html#in Nursing Competence Scale exhibited good reliability (average terpretation). inter-item correlation coefficients from 0.353–0.442, item-total Data were stored in a Microsoft Access 2000 database, and and Enterprise correlation coefficients from 0.322–0.731 and Cronbach’s alpha statistical analysis was performed using the from 0.79–0.91) and validity (content and current validity were programs studied). The Italian version of this tool shows adequate internal (MedCalc Software bvba, Ostend, Belgium; https://www.medcalc. consistency (>0.85) (Finotto & Cantarelli, 2009). The cluster of org; 2015). questions assessing teaching–coaching skills was not analysed all and MEDCALC Statistical R Software, SAS version 15.8 The level of significance adopted for the statistical tests was 5%. together; rather, we divided them into five groups (teaching the patient, teaching the family, teaching the student, evaluation of • the educational programme and teaching the professional team); 3.6 | Ethical considerations Diagnostic thinking: This outcome was measured using the Diag- The Local Ethics Committee approved the study protocol. All nurses nostic Thinking Inventory (Bordage, Grant, & Marsden, 1990), involved were informed about the study. Written informed consent which includes 41 items, each of which is self-rated by the nurse. was not considered a prerequisite for taking part in the study This tool evaluates flexibility in thinking (21 items) and the struc- because returning the questionnaires was already considered an ture of memory (20 items). The overall reliability of this tool was expression of consent. Written informed consent was also not 0.84 (alpha coefficient for internal consistency): 0.73 for flexibility required from inpatients because the information gathered was items and 0.75 for structure items. already part of their routine care. All inpatients were informed about PN using an information leaflet; unlike the nurses, the inpatients’ The following organisation-related outcomes were analysed: • consent was needed to collect data about their satisfaction with the care provided. Empowerment, which was analysed using the Empowering Leadership Questionnaire (ELQ) (Arnold et al., 2000; Bobbio, Manganelli Once stored in a password-protected database, the data were presented without any reference to individual nurses or patients. 1098 | DAL MOLIN ET AL. p = .0004). In addition, increases in thinking flexibility and the struc- 4 | RESULTS ture of memory were detected (from 92.39–96.34; p < .00001 and from 86.49–92.16; p < .00001, respectively). Based on Cohen’s sug- 4.1 | Nursing-sensitive patient outcome gestion, we found that the implementation of PN has a small effect Before the implementation of PN, data from 2,857 patients were or no effect in nursing competence (Cohen’s d effect size between collected and from 3,169 patients were collected after its implemen- 0.144–0.348), while it has a small or intermediate effect in diagnostic tation. thinking (Cohen’s d effect size between 0.473–0.571). As shown in Table 1, the numbers of patients with pressure The nurse’s leadership level increased in all five factors. How- ulcers (136–126; p = ns), falls (67–59; p = ns), urinary tract infections ever, not in all clusters of questions the differences were statistically (153–133; p = .0264) and venous catheter infections (peripheral: 61– significant and the analysis with Cohen’s d effect size indicated no 30; p = 0.0002—central: 12/215–3/295; p = .0080) decreased after effect. After the PN training, the team climate also improved in all the implementation of PN. factors. However, we find a small effect in two clusters of questions (vision of the group and task orientation). Data regarding staff- and organisation-related outcomes are pre- 4.2 | Staff- and organisation-related outcomes sented in Table 2. Data were collected from 369 nurses (82.38% female), and the rate of response for all questionnaires was higher than 85%. 4.3 | Satisfaction with care The nurses reported an increase in their competence; in particular, a statistically significant improvement was observed for the fol- Satisfaction with care was rated by 378 patients before PN imple- lowing competencies: “helping role” (18.11 vs. 19.85; p = .0001), mentation and by 315 patients after PN implementation. “diagnostic functions” (18.88 vs. 20.11; p = .0007), “managing situa- The satisfaction with nursing care increased, and according to tions” (27.60 vs. 28.60; p = .0210), “ensuring quality” (16.03 vs. Cohen’s suggestion, PN implementation has an intermediate effect in 16.93; p = .0057), “teaching-coaching the patient” (14.60 vs. 15.55; satisfaction of care (Table 3). p = .0002), “teaching-coaching the family” (5.60 vs. 5.98; p = .0015), Satisfaction with hospitalisation (not nursing care), which was “teaching-coaching the student” (4.72 vs. 5.14; p = .0335), “teaching- evaluated with a specific question, also showed improvement. How- coaching evaluation of the education programme” (8.29 vs. 8.90; ever, this change was not statistically significant (5.9 vs. 6.0; p = ns). T A B L E 1 Nursing-sensitive patient outcomes Preintervention (t0) N = 2,857 Postintervention (t1) N = 3,169 p Patient sample characteristics
Purchase answer to see full attachment