SEARCH

SEARCH BY CITATION

Keywords:

  • attrition;
  • new graduate;
  • New Zealand;
  • nursing workforce;
  • Registered Nurse;
  • separation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

Aim

To describe workforce separation rates and its relationship with demographic and work characteristics in the 2005 new graduate cohort's first 5 years as practising RNs in NZ.

Background

Retaining new graduate RNs is critical to nursing workforce sustainability; one study showed that if an RN is still employed in a hospital setting 5 years after graduation, he/she tends to remain active in the health industry.

Design

Retrospective analysis using the Nursing Council of New Zealand's registration data set for years 2005–2010.

Methods

All newly registered NZ graduates practising in NZ in 2005 (n = 1236) were tracked for 5 years.

Results

Within 5 years of graduation, 26% of the cohort had separated from the NZ nursing workforce, 18% in the first year. The under-25s (n = 517), 42% of the cohort, had the highest loss, 32%, in 5 years. Separations were significantly lower for graduates in their 30s vs. their 20s and for those who gained postgraduate tertiary qualifications post-registration (10%) vs. those who did not (29%). Hospitals were the most frequent employment setting over 5 years, the largest increase being community settings. Five-year retention rates in the four largest practice areas were surgical 26%, medical 16%, mental health 60% and continuing care 10%. After 5 years, 24% of those still practising (n = 920) worked in a different health board region.

Conclusions

New graduate RN losses were higher than in previous research, with younger RNs at most risk, threatening future sustainability of the nursing workforce and highlighting the need for evidence-based targeted strategies to retain them.

Why is this research or review needed?

  • The majority of research using large data sets has been conducted in the US; this study adds a New Zealand perspective, whose nurses actively participate in the global nursing market.
  • The study adds to descriptive analyses of new graduate attrition by determining relationships between separation, demographic and work characteristics.
  • Retention of new graduates in the nursing workforce is crucial for future nurse supply.

What are the key findings?

  • Five years after registration, 25·6% of the graduate cohort had left the NZ nursing workforce, with 18% leaving in the first year. Younger age was significantly associated with separation: the 5-year separation rate was 31·5% for the 20–24 years group that comprised 42% of the cohort. Ethnicity and gender characteristics were not significant.
  • Retention over 5 years was highest in mental health. Gaining a postgraduate tertiary qualification within 5 years of registration was associated with significantly lower separation.
  • Occupational mobility among those who did not separate was high across several domains, including employment setting, geographical area, urban/rural setting and clinical practice area.

How should the findings be used to influence policy/practice/research/education?

  • Separation rates for new graduate nurses differed by age group. Risks were highest for the youngest age group and lower for older groups, highlighting the importance of targeting retention strategies to sub-cohorts.
  • Aligning nursing education and new graduate transition programmes with primary care and community-based services would better prepare early-career nurses for practice in those settings.
  • Insights into what happens to early-career nurses and the significance of relationships between associated variables were determined, but many questions remain, best addressed in prospective longitudinal research.

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

Registered Nurse (RN) shortages have raised concerns in recent decades, threatening the sustainability of health service delivery and expected to worsen as older nurses retire and the gap between supply and demand widens (Buchan & Aiken 2008, Shacklock & Brunetto 2012). Common to developed countries, in the US, a downturn in numbers entering nursing schools and graduates being added to the register in the 1990s was eased in the 2000s by an upswing in younger entrants to nursing schools and workforces (Buerhaus et al. 2004, Auerbach et al. 2011). A substantial body of research (reviewed below) has found high attrition rates among recent RN graduates, highlighting the need to understand the factors that influence new graduates' career and attrition decisions (Sochalski 2002). In the case of New Zealand (NZ), where the study took place, entrenched shortages during the 1990s eased after a substantial nationwide pay increase in 2004 for nurses in the public health sector (Buchan & North 2009). Following the 2008 global financial crisis, nursing shortages declined further (Buerhaus et al. 2009, North 2010). Recruitment of younger people into nursing has improved in recent years. NZ new graduates and internationally qualified nurses were included in a longitudinal study employing descriptive statistics of all RNs added to the register in the 2005/6 year; after 5 years, 28% of the cohort was not practising, of whom some did not work in nursing from the first year (Nursing Council of New Zealand 2013). Indeed, an increasing number of recent graduates have been reported as not registering/entering the NZ nursing workforce (Ministry of Health 2009), which is threatened by dissociation, emigration and ageing. Despite NZ's remoteness, nurses are active participants in the global labour market: the number of nurses leaving for overseas (mainly Australia) almost doubled during 1992–2006 (Zurn & Dumont 2008, North 2010) and reliance on international recruitment increased, with overseas-trained nurses added annually to the register outnumbering NZ graduates (Cook 2009). Cook (2009) also noted that the proportion of NZ nurses under 35 years was at its lowest-ever level and that the new graduate population was older. With no compulsory retirement age, precise information on age[s] of retirement is unavailable. A large cohort of RNs is approaching pension-eligibility age (65 years): in 2011, 39% RNs were aged 50 and over, and 11% of the total were over 60 (Nursing Council of New Zealand 2012).

An understanding of early-career dynamics is needed to inform a strategic approach to retaining new graduates in the NZ nursing workforce. Based on a retrospective analysis of the administrative nurse registration data set maintained by the Nursing Council (the regulatory agency), this paper reports on what happened to the 2005 cohort of new graduate RNs in NZ in the 5 years after first registration and entry to the workforce. This study differs from other (mainly USA) retrospective analyses of existing data sets in several respects. It is one of a few to use administrative data sets (many use household labour surveys or other secondary large survey data).The entire annual output of newly registered RNs is included, not a sample. It differs from cross-sectional descriptions in that a cohort of new graduate RNs is followed up after 5 years. Although there is overlap between our cohort and that in the Nursing Council's longitudinal study, it differs by including only those RNs who were practising in nursing in NZ on registration, involved robust data verification and focuses on relationships between separation and demographic and work characteristics, additional qualifications and geographical location. The data set is recent, not historical, and reflects international issues (2008 global financial crisis) and local policies (pay increase, national implementation of a transition-to-practice programme) having an impact on separation and retention.

Background

Retrospective analyses of existing large data sets involving cohorts of RNs contribute to labour market research. In one of a series of analyses based on Current Population Surveys that documented US nurse labour market trends and predicted future supply (not attrition rates), Auerbach et al. (2011) reported for a RN sample of 68,611 that from 2002–2009, the number of new graduate RNs aged 23–26 years increased by 62%, a trend also noted by Buerhaus et al. (2004). A previous analysis of the 1973–2005 data (n = 86,568) found that RNs born in the mid-1970s had entered nursing later at an unprecedented rate, from two routes: previous careers and employment in other fields; and bachelor's degrees in other disciplines (Auerbach et al. 2007). Sochalski (2002) and Spetz et al. (2008) examined attrition using data from the 1990s–early 2000s. Using National Sample Survey data, Sochalski reported that 4·1% of female and 7·5% of male new graduates in 1996–1999 were not working in nursing after 1–4 years. Comparing the 1992/3 and 2000/1 cohorts of college graduates in California, Spetz et al. (2008) found that for the 2000/1 cohort, 1 and 3 years later, 89% and 88%, respectively, were working in health care; RN separation from the health industry was greater in the first few years after graduation among hospital-employed RNs, but if those nurses were still employed after 5 years, they tended to remain in the health industry. Retrospectively analysing 1991–2001 data from the British Household Panel Survey, Barron and West (2005) concluded that risk of separation increased with younger age, male gender, having a degree and being UK-born, but those who survived the first few years were likely to remain in nursing.

Characteristics of the nurse labour market have an impact on new graduate retention and turnover. A US survey found that 52·5% of a sample (n = 1765) worked within 40 miles of where they received secondary education (Kovner et al. 2011). In the UK, younger, childless new graduates were more mobile (Robinson et al. 2008). Uneven labour market demand means that some new graduate RNs cannot find jobs as nurses: Hirsch (2011) found that 15% of a survey sample of new graduates in the US was looking for nursing work. In Canada, employment of new graduate RNs increased from 59–71% over 4 years after a government-led initiative guaranteed a 6-month orientation-and-mentoring programme (Baumann et al. 2011). Cook (2009) and Crow et al. (2005) remarked on the high percentage of NZ and US RNs, respectively, working part-time; however, new nurses working full-time were more likely to leave in a Canadian study (Rhéaume et al. 2011).

Retaining RNs in the nursing profession in the first few years after graduation is crucial to nursing workforce sustainability (Crow et al. 2005, Zucker et al. 2006). New graduates not working in nursing (Finlayson et al. 2002, Sochalski 2002) worked in a wide range of fields in the US (Black et al. 2010) and Australia (Duffield et al. 2004), highlighting that nursing provided skills and qualities valued outside the nursing workforce. Some studies highlighted the influence of personal, workplace and organizational attributes and job opportunities on turnover intention and work behaviour (Kovner et al. 2007, 2009, Beecroft et al. 2008). Sochalski (2002) found that staff nurses (which would include new graduates) consistently are among the least-satisfied nurses. Issues influencing job satisfaction are regularly reported, including nurse shortages, workloads, poor person–job fit, workplace violence and bullying, shifts, pay, burnout and autonomy over work (Duffield et al. 2006, Black et al. 2010, Laschinger & Grau 2012).

These factors commonly underpin transition-to-practice interventions, that are found to reduce transition stress and turnover rates and to be cost-effective (Duchscher 2009, Dyess & Sherman 2009). Interventions include residency programmes (Sherrod et al. 2008), orientation and ‘on-boarding’ interventions (Beecroft et al. 2001, Friedman et al. 2011) and preceptors/mentoring (Beecroft et al. 2006, Zucker et al. 2006). Noting that the NZ health system is predominantly publicly funded and most acute hospital services are in the public sector, to increase retention of new graduates, since 2006, NZ public hospitals offered fully funded, structured 1-year programmes involving preceptors, professional development and structured learning, including formal postgraduate study. While not all new graduates find places in such programmes, 2006–2009 evaluations showed that participation improved retention (Haggerty et al. 2009).

The study

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

Aims

The aim is to describe 5-year workforce separation rates and the relationship with demographic and work characteristics for the new graduate cohort. We address the following questions regarding the cohort of new graduate RNs added to the register and working as nurses in NZ in 2005: What are the annual and 5-year separation rates of the cohort? What are the relationships between separation and demographic characteristics? What are the main geographical locations, employment settings and clinical practice areas in 2005 and for the remaining cohort in 2010 and relationships with separation? What is the relationship between additional qualifications and separation? Were there changes to weekly working hours reported in 2005 and for the remaining cohort in 2010?

Design

Retrospective cohort analysis using the Nursing Council of New Zealand's registration data set for years 2005–2010.

Sample and data

All NZ graduate RNs complete a 3-year bachelor's programme at a tertiary educational provider, then pass a compulsory State Registration Examination (offered twice each year), before applying to the Nursing Council to be added to the register. Each year RNs must apply for renewal of their Annual Practising Certificate (APC) at the time of their birthdays by providing evidence of competency and fitness to practise (Nursing Council of New Zealand 2008) and voluntarily complete a survey (~95% do so) to inform workforce policy. Anonymized data from APC application forms for 2005–2010 were supplied by the Ministry of Health. From this data set, we extracted the study cohort using these inclusion criteria: new to registered nursing; had no prior qualifications allowing registration as a nurse; graduated in NZ in the 2005 calendar year; and practising as a nurse in NZ. (We used the Nursing Council definition of ‘practising’ as ‘eligible to practise’ even if not currently active, e.g. looking for work.) The authors, in consultation with expert nurse professionals, manually screened the cohort to exclude any not meeting the selection criteria. This 2005 cohort (n = 1236), hereafter ‘the cohort’, was tracked for 5 years until 2010. The 2005 base year was selected for the following reasons. Prior to 2004 the register of nurses included midwives (a separate register was established for midwives in 2004) and any nurse registered to practise in NZ could renew her/his APC on payment of a small fee irrespective of whether they were practising. Following implementation of the Health Practitioners Competence Assurance Act in 2003, since 2004, APCs were contingent on competence to practise (active in a field of nursing and engaged in continuing education).

Ethical considerations

Formal ethical review for an anonymized administrative data set was not required.

Data analysis

Variables available in the data set found to be associated with separation in other studies were analysed. Independent sample and paired t-tests and one-way between-groups anovas were used to test for differences in continuous variables. The Shapiro–Wilk test was used to assess normality. When the assumptions for parametric tests were violated, their non-parametric equivalents were used. Chi-square tests were used to test for differences in categorical variables. Regression modelling was used to estimate the effects of demographic and work characteristics on the probability of those in the cohort separating from the workforce after 1 year. Backwards selection was used for the pre-specified predictors in the probit model. Interactions effects between key predictors were tested after initial estimation. All tests were two-tailed and the significance level was set at 5%. Stata SE v.11 was used for all statistical analyses.

Validity and reliability

Limitations of this secondary data set included: missing item response (the applicant did not respond to some questions), respondent error (the wrong code was entered by the applicant), data entry error (on the databases) and changes made over the period to the coded responses to the APC survey questions. Cross-validation using data from all other years was used to minimize errors in RN immutable attributes. High working hours (≥60 hours/week) were manually verified against other reported information in the same and adjacent years; if found to be erroneous it was imputed with values from the nearest year. Case-wise deletion was used for the analysis of all variables. Where missing data exceeded 1% in any question response, the percentage of valid responses is reported. Category labels were recoded so those that had changed (e.g. nursing practice area, employment setting and ethnicity where broad ethnicity categories were later specified) were consistent with the 2005 base year. Two of the authors (NN and WL) adjudicated on all data issues.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

In 2005 the cohort (n = 1236) was: mean age 30·3 years, range 20–68; 94% female; 75% European (including NZ and Other), 11% Asian, 7% Maori and 7% all other ethnicities.

Separation rates

After 1 year 82·2% of the cohort was practising as nurses. After 5 years, 74·4% of the cohort was practising in NZ; 25·6% (n = 316) had left the NZ nursing workforce. There was a loss of 17·8% (n = 220) in the first year (the largest annual separation rate of the 5 years) and annual attrition of 0·9%, 1·6%, 3·8% and 1·5% in years 2–5, respectively, until 25·6% was reached in 2010 (Figure 1). The 5-year separation rate for the 2006 cohort (the only other for whom data were available) was similar, 26·3%.

image

Figure 1. Cumulative separations from the NZ nursing workforce at end of years 2006–2010 of a cohort of new graduate RNs added to the register during 2005.

Download figure to PowerPoint

Relationships between demographic characteristics and separation

Age profile

The median age of the cohort in 2005 was 27 years. The modal age band was 20–24 years, but 32% were aged 35 or over. Five years later the median age of those still practising in NZ rose to 34 years (mean 36·1 years). The separation rate for the 20–24 years group (n = 517) comprising 42% of the cohort was steepest, losing 31·5% in the 5-year period and lowest for the 50–54 age band, no separations. Separation rates were highly significantly different between graduates in their 20s vs. graduates in their 30s (P = 0·007). There were no significant differences between those in their 30s vs. their 40s (P = 0·159). Figure 2 shows the 5-year separation rates for each 5-year age band (except 55 and over – numbers too small) of the cohort.

image

Figure 2. Five-year separation rates by age band.

Download figure to PowerPoint

Ethnicity and gender profile

Differences in 5-year separation rates amongst the three main ethnic groups, comprising European (75%), Asian (11%) and Maori (7%), were not significant (P = 0·235). Maori had the lowest separation rates at 21%, European 24% and Asians 30%. In 2005, females made up 94% of the cohort; there were only 74 males. Five-year separation rates for males were higher, 30%, compared with females, 25%, but not significantly so (P = 0·397).

Main geographical locations, employment settings and clinical practice areas in 2005–2010

Geographical location

APC applicants report on their employment location based on appended codes, each of which is classified as being an urban or rural environment. Response rate to this question was 93% (n = 1153) in 2005 and 99% (n = 911) in 2010. In 2005, 91% (n = 1044) worked in urban areas with 9% (n = 109) in rural environments. Differences in 5-year separation rates for urban (24%) vs. rural (32%) RNs were marginally insignificant (P = 0·067). For those still practising after 5 years, relocation to an urban/rural environment was highly significant (P < 0·001), with 7% (n = 54) of urban RNs moving to a rural area, but 55% (n = 41) of rural RNs relocating to an urban area. APC applicants also indicated the District Health Board (DHB) region where they worked, providing a second indicator of geographical mobility. DHBs (21 in 2005–2010) are public health organizations responsible for funding health services in their region and delivering public hospital and community services (Ministry of Health 2013a). In 2010, of those still practising in NZ, 76% continued to work in the same DHB region, but 24% were working in a different DHB region. Both indicators show substantial rates of geographical movement, implying high organizational turnover.

Employment settings

The main employment settings of the cohort were reported by 82% (n = 1011) of the sample in 2005 and 95% (n = 872) in 2010. Figure 3 shows that hospital settings (predominantly public hospitals) were the main employment setting in 2005 (73%) and 2010 (63%). Only community employment settings (e.g. primary health/general practice and public health) substantially increased their share, from 6% of the cohort in 2005 to over 26% 5 years later. After 5 years, 54% of the 2005 hospital-based cohort (n = 739) still worked in hospital settings, 16% worked in community settings and 21% had separated; the remaining 9% worked in other settings or did not respond.

image

Figure 3. Employment setting.

Download figure to PowerPoint

Areas of nursing practice, change and relationship with age

The response rate to ‘main practice area’ question was 93% (n = 1151) in 2005 and 95% (n = 871) in 2010. Figure 4 charts the four largest areas of nursing practice for the cohort in 2005 and the four areas of nursing practice showing the most growth in numbers from that cohort 5 years after registration. These eight areas of nursing practice account for 65% (n = 778) and 78% (n = 687) of valid responses in 2005–2010 respectively.

image

Figure 4. Selected practice areas.

Download figure to PowerPoint

In 2005, the four areas of practice with the highest number of new graduate RNs (in descending order) were surgical, medical, mental health and continuing care (i.e. older persons' services, mainly private residential care). The RNs who listed one of these four areas as their main area of practice in 2005 were followed up in 2010 to track change to clinical area (Table 1). There was high separation from these areas with the exception of mental health (P < 0·001). There were highly significant differences in age between those remaining in the medical practice setting and those leaving (P = 0·004), with those leaving being younger, 29·4 years old (median 25 years), than those remaining, 36 years old (median 37 years).

Table 1. Four largest sub-cohorts by practice areas in 2005.
Sub-cohort practice areaPractising in 2005Practice area in 2010
AnySameRelated
  1. a

    Related areas are emergency & trauma and peri-operative care.

  2. b

    Related areas are assessment and rehabilitation.

Surgical206153 (74%)53 (26%)33 (16%)a
Medical180130 (72%)28 (16%) 
Mental health120105 (88%)72 (60%) 
Continuing care11287 (78%)11 (10%)10 (9%)b

Additional qualifications gained and relationship with separation

Postgraduate certificates/diplomas gained by the cohort within 5 years post-registration were analysed. Excluding vocational certification, 232 RNs gained a further qualification within 5 years of registration; 209 of these were still working in 2010. By 2010, 23% of the remaining cohort (n = 209) held a further qualification. There was a highly significant difference (p < 0·001) in 5-year separation rates between those gaining further qualifications, 10%, vs. those not, 29%. The proportion of RNs with postgraduate tertiary qualifications amongst the eight clinical practice areas described above was significantly different (P < 0·001), with 65% of the graduate cohort working in mental health in 2010 having gained a postgraduate tertiary qualification.

Working hours

Total weekly working hours of the cohort in 2005 and of those still practising in NZ in 2010 were analysed by age band. RNs not reporting any working hours in 2005 (24%, n = 299), mostly those not in paid employment, were excluded. The mean weekly hours, for those who reported non-zero working hours, were 35·7 (sd 7) and 34·5 (sd 8) in 2005 and 2010 respectively. There were no significant differences in reported working hours amongst age bands in 2005 (Kruskal–Wallis, P = 0·328) or 2010 (Kruskal–Wallis, P = 0·623). For those reporting working hours in 2005, those in the following age bands reported significantly less working hours per week 5 years later (Wilcoxon signed-rank, 0·001 < P < 0·027): age 20–24, 4·2 less hours; age 25–29, 4·9 less hours; age 30–34, 4 less hours; and age 45–49, 3·5 less hours.

Regression model

The Hosmer–Lemeshow tests for the estimated models were highly significant, indicating poor goodness of fit. Despite using a national cohort, low numbers outside the main clinical practice areas and the few, for instance, rural or male RNs led to many covariate patterns with low numbers and even fewer, if any, events (separations). This is compounded by the large number seeking positions in their first year, further reducing the effective sample size. Omitted-variable bias may have occurred as the APC data set is highly limited; pertinent information, such as job satisfaction, own/partner health status, intention to leave and household income, probably will have improved model fit. Both factors could have caused models to have unstable parameter estimates; therefore, the results have not been reported.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

A key result is that more than a quarter of the cohort had separated from the NZ nursing workforce 5 years later, with 18% separating in the first year. Compared with other retrospective data set analyses, the separation rate is much higher than in the US, where rates of 4·1%, 4 years later (Sochalski 2002), and 11% and 12%, 1 and 3 years later (Spetz et al. 2008), were reported. Both Spetz et al. (2008) and Barron and West (2005) concluded that nurses who survive the first few years after transitioning to the workforce were more likely to remain in nursing.

Younger age was the only demographic factor associated with separation. The age distribution of the cohort shows that, as in the US (Buerhaus et al. 2004, Auerbach et al. 2011), nursing is attracting younger people, but the challenge is to retain them in the NZ workforce. New graduates under 25 years of age, comprising over four-tenths of the cohort, were most at risk of early separation – 32% in 5 years. In a Canadian survey, where 83% were under 30 years (Rhéaume et al. 2011), younger age was not associated with intent to leave. The cohort was predominantly female, as in the US and Canada, where Kovner et al. (2007) and Rhéaume et al. (2011) reported that 80% and 95%, respectively, of their new graduate samples were female. Differences in male and female separation rates for the cohort were insignificant, unlike in the UK (Barron & West 2005) and US (Black et al. 2010), where male gender was associated with higher separation. Ethnicity was not associated with separation rates. As in the US, where Kovner et al. (2007) observed European domination of new graduate RNs, in NZ, ethnicity and gender profiles of the cohort did not reflect the general population (in 2006 European 68%, Maori 15%, Asian 9%, Pacific 7%) (Ministry of Health 2013b, Statistics NZ 2013). Sochalski (2002) found that rates for males not working in nursing were higher than for females at 1–4 and 5–8 years after graduation for the 1992–1995 and 1996–1999 graduates (but not for 1988–1991 graduates) and that males at all ages were less satisfied with nursing than females regardless of position, employment setting and years after graduation.

Nearly three-quarters of the study cohort worked in hospitals in the first year of practice and after 5 years, nearly two-thirds still did so. If observations of Spetz et al. (2008) hold true, that nurses working in hospitals 5 years after graduation tend to stay in the industry, prospects for future retention of half the sub-cohort are positive. Reported rates for new graduates working in hospitals were higher in Canada at 92% (Rhéaume et al. 2011) and in the US where rates of 82–89% have been reported (Kovner et al. 2007, 2011, Spetz et al. 2008). In 5 years, community practice settings increased their share from 6–26% of the cohort, higher than in the US where after 3 years (2000/1 cohort), 42% of those still nursing, but no longer in hospitals (3% of the cohort), had moved to ambulatory settings (Spetz et al. 2008). Reflecting the strengthening of primary health care in policy since 2001 (Tenbensel et al. 2008), a smaller proportion of NZ new graduates worked in hospitals and a higher proportion in community settings than have been reported elsewhere.

Similar to Kovner et al. (2007), our results indicate that in their first years the new graduate cohort is an occupationally mobile group. This is evident in the separation rates from the NZ nursing workforce, the change from hospital to community settings and other high rates of movement in employment setting, practice area and DHB region.

Other indications of turnover are the high rates of change by 2010 for the top four practice areas in 2005. High turnover of nurses, including new graduates, from the primary work unit has been found in another NZ study where new graduates comprised over 40% of new recruits (North et al. 2013) and in the US (Beecroft et al. 2001, Bowles & Candela 2005, Zucker et al. 2006). While these studies are based on unit- or organizational-level studies, our results illustrate workforce turnover in a 5-year period for a country's entire RN new graduate cohort. In contrast with Kovner et al. (2011), who found that new graduate nurses in the USA tended to have low geographical mobility, urban-rural mobility and regional mobility (as RNs moved to a different DHB) were high in our cohort.

For the NZ cohort, occupational mobility and that nearly a quarter of the cohort in 2010 had gained postgraduate qualifications may be related to participation (by some, not all) in the year-long transition-to-practice programmes offered nationwide, mainly in public hospitals, from about the time the cohort entered the workforce. In hospital settings, surgical and medical were the top two practice areas, possibly reflecting the preference of these two areas in programmes (Haggerty et al. 2009). On completing their year in the clinical area[s] where the programme is offered, new graduates will be clearer about their preferred clinical area or place. For others, occupational mobility probably reflects their not initially getting a position in their preferred area. An example is seen in the drop-off of new graduates in continuing care with only 11 of 112 (in 2005) still in the same practice area 5 years later; new graduate recruitment data in 2012 showed that only 7 (of 1232) new graduates nominated aged residential care as their preferred practice area (Ministry of Health 2013b). Working full-time was related to leaving in Canada (Rhéaume et al. 2011); the decline in average weekly working hours after the first year may reduce separation. It may also reflect participation in the full-time transition-to-practice programmes in the first year. That the change after 5 years is significant only for age bands 20–34 and 45–49 suggests that the nurse is balancing work with family needs (Crow et al. 2005).

We found no previous studies reporting that gaining a postgraduate qualification was associated with significantly lower separation. A third of our cohort with postgraduate qualifications worked in mental health, the area with highest retention. The finding could infer a commitment to nursing as a career. It also reflects the inclusion in many transition programmes of postgraduate academic study (Haggerty et al. 2009). A small survey involving Australian and NZ new graduates found the NZ group more likely to engage in postgraduate study (Huntington et al. 2012). The level of pre-registration qualification and its association with separation has been reported in other research, e.g. in the UK, where higher qualifications (whether or not the nurse has a degree) were shown to be a risk factor for separation (Barron & West 2005) and the US where RNs with a degree are more likely to be working outside nursing (Black et al. 2010). In NZ, all new RNs graduate with a bachelor's degree.

Limitations

A limitation of the present study lies in its retrospective analysis of a data set involving responses to the questions in the Nursing Council's Application for Practising Certificate form. Additional information is needed on a range of factors known to be associated with separation, turnover and retention, such as job satisfaction, satisfaction with nursing, health, turnover and separation intentions and reasons for not registering and engaging with the NZ nursing workforce including intention to emigrate. The lack of information on and adjustment for known separation determinants suggests that confounding cannot be excluded as an explanation for the reported associations.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

Although more and younger RNs are being educated in NZ and added to the register compared with a decade previously, a substantial proportion are not being retained in the NZ nursing workforce. The high separation rates, particularly of younger nurses and after the first year of practice, raise questions about how young people are recruited to nursing, their learning and clinical experiences as students. How is nursing as a career marketed to potential students? How well prepared are nursing students for the realities of practice? Are effective programmes to develop personal strategies, e.g. resilience and stress management, offered?

This analysis adds a workforce dimension to studies measuring new graduate RN turnover at unit and organizational level, by observing a national cohort of all newly graduated RNs. Implications of the high occupational mobility of new graduates are that student nurses need to be exposed to the clinical settings they are likely to work in as graduates. In particular, new graduates who first worked in hospitals and who moved were most likely to move to community settings and primary health care, reflecting NZ's primary health strategy. It is important that student nurses are exposed to community-based services and primary health care and that these services offer transition-to-practice programmes.

High early-separation rates of the youngest new graduates occurred when two measures associated with retention (substantially increased pay and fully funded transition-to-practice programmes) had been recently implemented for DHB-employed nurses. This finding highlights the need for improved workforce intelligence on RNs' early-career decision-making. Evaluations are needed of the impacts of recent NZ initiatives (Nursing Council's cohort study, the Ministry of Health's system to facilitate new graduate recruitment and transition-to-practice programmes) on new graduate retention and turnover. In light of global competition for nurses and the higher pay offered by competing countries, an understanding of nurses' decisions to leave nursing/NZ is essential.

The analysis has provided insights into what happens to early-career RNs and has determined the significance of relationships between associated variables, but leaves many questions unanswered. For example, do separated RNs subsequently rejoin the workforce? A future retrospective analysis of the cohort, e.g. after another 5 years, could provide answers. The small size of the cohort and that the Nursing Council's Application for Practising Certificate form does not collect some data (e.g. job satisfaction, household income, partner status) limited our ability to identify predictors of separation risk. Adding subsequent new graduate cohorts for whom there are 5 years of data would address the limitation of size and supplementing the present questionnaire would reduce variable limitations. The administrative data set does not include data on new graduates' reasons for early separation and occupational movement, questions best answered by prospective cohort designs, not cross-sectional and retrospective designs. A longitudinal study has the potential to produce insights into new graduate career decisions, allowing interventions to retain new graduate nurses to be tailored to those nurses at highest risk and to target critical periods early in the nursing career.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

We thank the Nursing Council of New Zealand for permission to use the data, clarifying issues for us and commenting on a presentation of our provisional results on June 2012. The Ministry of Health supplied the raw data used for this analysis.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References

All authors have agreed on the final version and meet at least one of the following criteria [recommended by the ICMJE (http://www.icmje.org/ethical_1author.html)]:

  • substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data;
  • drafting the article or revising it critically for important intellectual content.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. The study
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. Funding
  10. Conflict of interest
  11. Author contributions
  12. References