Low job satisfaction does not identify nurses at risk of future sickness absence: Results from a Norwegian cohort study

Abstract

Background

Sickness absence is high in healthcare and contributes to nursing staff shortages reducing the efficiency and quality of patient care. Assessing the risk of sickness absence in working nurses opens opportunities for preventive strategies. Job satisfaction has attracted much attention in healthcare research and has been associated with sickness absence among nurses.

Objectives

To investigate if job satisfaction scores are useful to identify working nurses at risk of future sickness absence.

Design

Prospective cohort study with a baseline period from November 2008 to March 2009 and 1-year follow-up.

Settings

Hospitals, nursing homes, and ambulant care settings in Norway.

Participants

2059 Norwegian nurses, of whom 1582 (77%) could be followed-up.

Methods

Nurses received a questionnaire at baseline and after 1-year follow-up. The questionnaire contained the Job Satisfaction Index (JSI), a 5-item scale measuring overall job satisfaction, and asked for sickness absence in the last 12 months. Baseline JSI scores were included in a logistic regression model with self-rated sickness absence at 1-year follow-up as outcome variable. Predictions of sickness absence were calibrated by the Hosmer–Lemeshow goodness-of-fit test. The ability of JSI scores to discriminate between nurses with and without sickness absence was examined by receiver operating characteristic analysis and expressed as area under the curve (AUC).

Results

Low job satisfaction was associated with higher odds of sickness absence (odds ratio [OR] = 1.05; 95% confidence interval [CI] 1.01–1.09) and high (≥31 days) sickness absence (OR = 1.10; 95% CI 1.06–1.14). Calibration was acceptable, but job satisfaction neither discriminated between nurses with and without sickness absence (AUC = 0.54; 95% CI 0.51–0.58) nor between nurses with and without high (≥31 days) sickness absence (AUC = 0.58; 95% CI 0.54–0.63).

Conclusions

The results of this study indicated that job satisfaction was associated with sickness absence, though job satisfaction scores as measured with the JSI did not identify working nurses at risk of sickness absence.

Introduction

Nursing shortages are a major problem in many countries, where the demand for nursing care is increasing and nursing responsibilities are widening. The shortage of registered nurses will worsen, because demands for nursing services are on the increase due to longer life expectancies and higher numbers of people living with chronic disease. In the United States, the demand for nursing services will exceed the supply by nearly 30% in 2020 (Andrews and Dziegielewski, 2005). The ageing workforce, declining enrollment of new nurses, and high turnover intentions have been identified as the main causes for current nursing shortages (Janizewski Goodin, 2003, Coomber and Barriball, 2007, Duvall and Andrews, 2010).

The nursing staff is further reduced by high sickness absence levels in healthcare (National Survey of the Work and Health of Nurses, 2005, Occupational Safety and Health Administration, 2010, National Health Service, 2011), resulting in increased work pressure and lower morale of the remaining staff (Aiken et al., 2002, Lang et al., 2004). Consequently, the efficiency and quality of nursing care decrease, which adversely affects patient outcomes (Kane et al., 2007, Aiken et al., 2012). In a systematic review of the literature, Davey et al. (2009) found that an individual nurse's work attitudes were strongly associated with sickness absence. Work attitudes refer to personal feelings or beliefs associated with work, such as job and work involvement, organizational commitment, group commitment, occupational commitment, and job satisfaction.

Job satisfaction has attracted much attention in healthcare research. There is a large body of knowledge about determinants of job satisfaction among nurses. From a systematic review of the literature, Lu et al. (2005) reported that work stress, organizational commitment, depression, and cohesion within a nursing team were most strongly correlated with job satisfaction of nurses, with Pearson correlation coefficients (r) >0.50. Moderate associations (0.20 < r < 0.50) were found for autonomy, supervisor and co-worker support, and both collaboration and communication with medical staff. Personal factors such as age, professionalism, years of experience, and job involvement were weakly associated (r < 0.20) with job satisfaction among nurses. A meta-analysis of 31 studies representing a total of 14,567 nurses in various healthcare settings showed that job satisfaction was most consistently correlated with work stress, followed by nurse–physician collaboration, and autonomy (Zangaro and Soeken, 2007). Although a single factor at a given point in time may lead a nurse to consider a job satisfying or not, job satisfaction is generally multifactorial, i.e. determined by a combination of intra-personal, inter-personal, and work factors (Utriainen and Kingäs, 2009, Hayes et al., 2010, Lu et al., 2012).

Most studies in the eighties and early nineties failed to find a significant relationship between job satisfaction and sickness absence among nurses. Taunton et al. (1995) studied 1107 nurses working in four U.S. hospitals and reported that sickness absence decreased if job satisfaction increased. However, the authors used an untitled non-validated job satisfaction instrument. Matrunola (1996) developed a job satisfaction instrument adjusted to the setting of a British district hospital and found no significant association between job satisfaction and sickness absence among 50 nurses of an elderly care unit. It should be acknowledged that the author investigated a very small and non-representative sample of nurses with a non-validated job satisfaction instrument.

Gauci Borda and Norman (1997) studied job satisfaction in a sample of 254 Maltese hospital nurses with the Global Satisfaction Index, which is a validated instrument to measure overall job satisfaction. The authors reported that overall job satisfaction was negatively associated with the recorded frequency of short-term sickness absence, but not with the number of sickness absence episodes lasting longer than three days. However, they measured job satisfaction in February 1996 and sickness absence in the period February 1995 to February 1996. Thus, sickness absence preceded the measurement of job satisfaction and, therefore, we cannot infer that the higher frequency of short-term sickness absence was the result of low job satisfaction.

Song et al. (1997) annually measured job satisfaction with the Job Diagnostic Survey in 143 South Korean hospital nurses during 4 years. Overall job satisfaction did not differ between nurses working in a special care unit (N = 34) and nurses working in the intensive care unit (N = 109), though satisfaction with payment and supervision was higher in the special care unit. Sickness absence levels were lower in the special care unit (1.2%) than in the intensive care unit (2.8%). The authors concluded that higher satisfaction and lower sickness absence rates coincided in the special care unit. This may be true for specific aspects of satisfaction, but not for overall job satisfaction which did not differ between units. Furthermore, the results are difficult to interpret, because job satisfaction and sickness absence were investigated at unit level, not at the individual level.

Siu (2002) studied overall job satisfaction at the individual level in two samples of Hong Kong hospital nurses. In one sample (N = 144), job satisfaction was significantly related to the number of self-reported sickness absence days during 1-year follow-up, while in the other sample (N = 114) no significant associations were found. These inconsistent findings may be explained by unmatched gender ratios and nursing settings as there were more men (33%) and more psychiatric nurses (49%) in the larger sample, in which job satisfaction was associated with sickness absence, as compared to 15% and 25%, respectively in the other sample.

The results from previous studies are inconclusive, so we need more studies to investigate the relationship between job satisfaction and sickness absence in nurses. Better understanding of associations between job satisfaction and sickness absence is important to assist nurse managers in preventing sickness absence and improving nurse staffing. This study investigated the prospective associations of job satisfaction with sickness absence among Norwegian nurses.

We specifically wanted to know if job satisfaction scores forecast the risk of sickness absence and identify working nurses at risk of sickness absence. For that purpose, prospective associations between job satisfaction and sickness absence are not sufficient. An association neither tells us if job satisfaction scores accurately predict future sickness absence (calibration), nor if job satisfaction scores discriminate individuals with sickness absence from those without sickness absence (discrimination). This is the first study that addresses the calibration and discrimination characteristics of job satisfaction as predictor of future sickness absence in nurses. If calibration and discrimination are adequate, then job satisfaction scores can be used to identify working nurses at risk of sickness absence and invite them for preventive counseling.