Sickness absence in student nursing assistants following a preventive intervention programme

Transcription

1 SHORT REPORT Occupational Medicine 2011;61:57 61 Advance Access publication on 13 September 2010 doi: /occmed/kqq Sickness absence in student nursing assistants following a preventive intervention programme A. L. Svensson 1, J. L. Marott 2, P. Suadicani 1, O. S. Mortensen 1 and N. E. Ebbehøj 1 1 Department of Occupational and Environmental Medicine, Bispebjerg University Hospital, Copenhagen 2400 KBH NV, Denmark, 2 The Copenhagen City Heart Study, Bispebjerg University Hospital, Copenhagen 2400 KBH NV, Denmark. Correspondence to: A. L. Svensson, Department of Occupational and Environmental Medicine, Bispebjerg University Hospital, Copenhagen 2400 KBH NV, Denmark. Tel: ; fax: ; lyng.annemarie@gmail.com Background We have previously shown that a multidimensional programme combining physical training, patient transfer techniques and stress management significantly reduced sickness absence rates in student nurse assistants (NAs) after 14 months of follow-up. At follow-up, the control group had reduced SF-36 scores for general health perception [general health (GH)], psychological well-being [mental health (MH)] and energy/fatigue [vitality (VT)] compared with the intervention group, which remained at the baseline level for all three measures. Aims To ascertain whether this effect remained after a further 36 months of follow-up and to analyse the association of GH, MH and VT scores with sickness absence. Methods This was a cluster randomized prospective study. The original study involved assessment at baseline and follow-up at 14 months (the duration of the student NA course). Of 568 subjects from the original intervention study, 306 (54%) completed a postal questionnaire at 36 months. Results Sickness absence increased in both groups between the first and second follow-up. At the second follow-up, the intervention group had a mean of 18 days of sickness absence compared with 25 in the control group but this was not significant. GH at 14 months follow-up was found to predict sickness absence levels after 3 years. MH and VT scores showed an inverse association with sickness absence but the results were not significant. Conclusions The results suggest that the initial intervention did not have a sustained effect on sickness absence 36 months after initial follow-up of the study group. Key words Assistant nurses; sickness absence; intervention study. Introduction In Denmark, working days equivalent to fulltime employees (5% of the total workforce) are lost every year due to sickness absence [1]. Sickness absence among health care workers is a substantial problem [1]. Despite this, few studies of prevention of sickness absence among health care workers have taken place [2 3]. A recent review on work health promotion and sickness absence by Kuoppala et al. [4] showed that education on exercise, lifestyle and ergonomics might be effective in preventing sickness absence. We have previously shown that student nurse assistants (NAs) allocated to a multidimensional intervention programme aiming to reduce low back pain (LBP) had lower rates of sickness absence compared with a control group after 14 months of follow-up [5]. The intervention programme consisted of three preventive measures: physical training, patient transfer techniques and stress management [5]. In the control group, lower scores on general health perception [general health (GH)], energy/fatigue [vitality (VT)] and psychological well-being [mental health (MH)], all items from SF-36, were observed after 14 months, whereas in the intervention group, these measures remained at the baseline level [5]. The intervention effect on GH, VTand MH scores suggests that the beneficial effect of the intervention on sickness absence may be due to its effect on these SF-36 health measures. The aim of the study was to investigate if the observed intervention effect on sickness absence was a brief one or was still present after a further 3 years, and furthermore to analyse the association of scores of GH perception, energy/fatigue, and MH with sickness absence, Ó The Author Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please journals.permissions@oup.com

2 58 OCCUPATIONAL MEDICINE with the objective of better understanding mechanisms by which the intervention effect may operate, so that future interventions might be tailored to reduce sickness absence more efficiently. Methods The present study is a follow-up of a cluster randomized prospective study. The original study ( ) involved a baseline assessment in the first week of the first semester and follow-up at 14 months (a week before the last exam on completion of the students training) and has been described in detail elsewhere [5,6]. A postal questionnaire was sent to the participants 36 months after the first follow-up. All participants gave written informed consent. Ethical approval was granted by the Copenhagen Ethics Committee. The intervention consisted of an integrated programme of three preventive measures, physical training (48 h), patient transfer techniques (20 h) and stress management and personal development (22 h). All intervention group members were offered the same generic intervention. The control group followed the standard programme with no additional training. A description of the intervention programme has previously been published [5]. Sickness absence was self-reported, in response to the question How many days during the last 12 months have you been absent due to your own sickness? [7]. GH perception, VT and MH were measured by SF-36 scales, with higher scores representing better health [8]. Differences in baseline characteristics between the intervention and the control group were tested with Fisher s exact test and Student s t-test, as appropriate. The effect of the intervention on outcome measures at follow-up was analysed according to the intention to treat principle, regardless of whether or not subjects actually participated in or adhered to the intervention. The analysis was conducted with all available respondents at the time of the second follow-up. The effects of the intervention on the health measures listed in Table 2 were evaluated with mixed-effects models with intervention and time as well as their interaction as fixed effects, taking into account the random variation between students and residual variance between single measurements. The association of SF-36 measures with sickness absence was analysed with mixed-effect models and logistic regression analysis. All mixed-effects models were adjusted for age and intervention (SAS, version 9.1, procedure Mixed). Results At baseline, 906 NA students were invited to participate in the study and 790 (87%) completed the questionnaire. The response rate at the first follow-up was 584/790 (74%) and at the second follow-up it was 306/790 (39%). Of 329 intervention subjects who completed the first follow-up questionnaire, 177 (54%) completed the second follow-up questionnaire; in the control group, the figures were 255 and 129 (50%), respectively. There was no difference in baseline measures between the intervention and the control groups, as shown in Table 1. Non-responders in the intervention group were younger than responders (mean age 32, SD 9) 10 versus 36, SD 12, P, 0.05, while there were no age differences between responders and non-responders in the control group. There was no difference between responders and non-responders regarding sickness absence, gender and prevalence of LBP present continuously for 3 months (first follow-up data). Sickness absence (mean number of days reported in the preceding 12 months) increased in both groups between the first and the second follow-up assessments. At the second follow-up, the intervention group reported a mean of 18 days sickness absence compared with 25 in the control group (not statistically significant) (Table 2). The estimated effect difference was 7.5 (95% CI to 1.0), i.e. non-significant. Sickness absence was further log-transformed due to its positive skew still yielding a positive but non-significant intervention effect. The intervention group reported fewer episodes of sickness absence of.10 days duration, OR (95% CI ), but this was also non-significant. At the second follow-up, the intervention group had a mean GH score of 74 versus 70 (P, 0.05) in the control group, with an estimated effect difference of 4.1 (95% CI ). Table 1. Characteristics of the study population (student NAs, N 5 306) and baseline values of the outcome measures Intervention N (58%) Control N (42%) Mean (SD) Mean (SD) Sickness absence (days) 4 (8) 4 (7) GH a 81 (15) 80 (15) MH a 79 (15) 80 (14) VT a 71 (17) 69 (18) Females N 160 (90) 108 (84) Age 34 (12) 35 (11) Body mass index 25 (5) 24 (4) Episodes of LBP 0.13 (0.7) 0.18 (1.0) N (%) N (%) LBP for continuously 3 12 (7) 8 (6) months Leisure time physical 71 (41) 50 (39) activity Smoking 82 (46) 48 (37) Country of birth, 132 (76) 95 (76) Denmark Education,9 years of schooling 35 (20) 21 (17) Fisher s exact test and Student s t-test was used when appropriate. a Range:

3 A. L. SVENSSON ET AL.: SICKNESS ABSENCE IN STUDENT NURSING ASSISTANTS 59 Over the entire study period, both groups had a decline in GH, VT and MH but this was most pronounced in the control group. Continuous LBP for.3 months increased equally in both groups throughout the study period. Table 3 shows the association between GH, VT and MH at 14 months follow-up and sickness absence at the second follow-up. Scores for each were inversely associated with sickness absence, although only GH showed a significant association (estimated effect difference 0.51, 95% CI to 20.12, P, 0.05). In a Table 2. Outcome measures at 14 months follow-up and 3 years later in the intervention and control groups and the estimated effect of the intervention Intervention group (N 5 177) (58%) Mean (SD) Control group (N 5 129) (42%) Mean (SD) Estimated effect a (P value) 95% CI Sickness absence during the last 12 months (days) 12 (21) 18 (38) 26.3 NS to 2.1 At 3 years follow-up 18 (52) 25 (59) 27.5 NS to 1.0 GH b 80 (16) 77(16) 3.3 NS 20.5 to 7.1 At 3 years follow-up 74 (18) 70 (19) 4.1* 0.3 to 7.9 VT b 68 (17) 63 (19) 4.9** 1.1 to 8.7 At 3 years follow-up 59 (15) 57 (17) 2.8 NS 21.0 to 6.6 MH b 77 (15) 76 (17) 1.4 NS 21.8 to 4.7 At 3 years follow-up 68 (12) 68 (13) 0.1 NS 23.1 to 3.4 LBP for continuously 3 Frequency (%) Frequency (%) OR c months 22 (12) 13 (10) 1.5 NS 0.5 to 4.4 At 3 years follow-up 67 (39) 51 (40) 0.9 NS 0.6 to 1.5 Sickness absence.10 days during the last 12 months 54 (32) 51 (41) 0.7 NS 0.4 to 1.1 At 3 years follow-up 41 (25) 38 (31) 0.7 NS 0.4 to 1.2 a Effect difference was adjusted for age and baseline data. b Range: c Effect difference was adjusted for age and baseline value. P values for the mixed-effect model: NS, not statistically significant, P. 0.05, *P, 0.05, **P, Table 3. SF-36 measures at 14 months follow-up and the association with sickness absence after 3 years further follow-up No. of participants Estimate a 95% CI P value GH b (20.89 to 20.12) ** VT b (20.53 to 0.25) NS MH b (20.84 to 0.51) * SF-36 measures at 14 months follow-up and the association with.10 days of sickness absence at 3-years further follow-up OR c 95% CI P value GH b 0.98 (0.97 to 0.99) NS VT b 1.01 (0.99 to 1.03) * MH b (0.98 to 1.02) NS a Mixed-effect model adjusted for age and intervention. b Range: c Logistic regression analysis adjusted for age and intervention. NS, not statistically significant, P. 0.05, *P, 0.05, **P, 0.01.

4 60 OCCUPATIONAL MEDICINE logistic regression model, GH was also inversely associated with episodes of.10 days of sickness absence at the second follow-up (OR , 95% CI , P, 0.05). There was no significant difference between the intervention and the control group in respect of their participation in the labour market, with 65% working in the eldercare sector, 11% employed in other sectors, 13% taking other education courses and 11% not active in the labour market. Discussion Sickness absence increased in both groups between the first and the second follow-up assessments. The intervention group had a lower level of sickness absence compared with the control group, but this was not statistically significant. The estimated effect difference on sickness absence between the groups at the second follow-up was 7.5 days. A smaller but statistically significant difference in reported sickness absence between the intervention and the control groups (5.9 days) was found at the first follow-up [5]. These results do not suggest a sustained intervention effect on sickness absence at second follow-up 3 years later. The intervention group had a higher GH score compared with the control group at the second follow-up, although between the first and the second follow-up both groups showed a decline on the SF-36 scales of GH perception, VT and MH, indicating deterioration in health after the NA students finished their education. In this study, GH scores at 14 months follow-up were inversely associated with reported sickness absence at the second follow-up. MH and VT scores also showed an inverse association with sickness absence but these results were not significant. The GH scale is a strong independent predictor of morbidity and mortality [9 13]. GH, as assessed by one single item from the GH scale, How do you rate your health in general?, with five response options (very good, good, fair, poor and very poor) has previously been used as a predictor of sickness absence [14 15]. We found no difference between the groups with respect to involvement in the labour market. The high sickness absence rates reported by both groups at the second follow-up may be a result of the observed deterioration in SF-36 health scores, although there are also behavioural and cultural determinants of sickness absence, which may represent an aspect of an individual s response to difficulties related to the workplace or their personal life [14]. In some workplace cultures, a degree of sickness absence may even be regarded as an employee benefit contributing to higher satisfaction with psychosocial work conditions [15]. Further studies evaluating interventions to prevent sickness absence should focus ways of preventing deterioration in health indices and facilitating efficient coping strategies. In order to confirm levels of sickness absence more objectively, they should also wherever possible have access to employers sickness absence records. Key points Reported levels of sickness absence had increased in both the intervention and control groups at the second follow-up 3 years later in this study. Lower general health scores at 14 months followup were associated with higher reported levels of sickness absence at follow-up 3 years later. Future studies focussing on interventions to prevent sickness absence should focus on preventing deterioration in health indices such as general health perception, vitality and psychological well-being. Funding The Danish Working Environment Research Fund, The Danish Ministry of Employment. Conflicts of interest None declared. References 1. Danish Ministry of Employment. Analyse af sygefraværet (20 August 2009, date last accessed). 2. Tveito T, Eriksen HR. Integrated health programme: a workplace randomized controlled trial. Journal of Advanced Nursing 2009;65: Brox JI, Frøystein O. Health-related quality of life and sickness absence in community nursing home employees: randomized controlled trial of physical exercise. Occup Med (Lond) 2005;55: Kuoppala J, Lamminpää A, Husman P. Work health promotion, job well-being and sickness absences a systematic review and meta-analysis. J Occup Environ Med 2008;50: Svensson AL, Strøyer J, Ebbehøj N et al. Multidimensional intervention and sickness absence in assistant nursing students. Occup Med (Lond) 2009;59: Svensson AL, Støyer J, Ebbehøj NE, Mortensen OS. Factors predicting dropout in student nursing assistants. Occup Med (Lond) 2008;58: The National Research Centre for the Working Environment. B8rgeskemaer.aspx?lang5da (20 November 2009, date last accessed). 8. Ware JE, Kosinski M. SF-36 Physical & Mental Health Summary Scales: A Manual for Users of version 1 (2nd edn). Lincoln, RI: QualityMetric, 2001.

5 A. L. SVENSSON ET AL.: SICKNESS ABSENCE IN STUDENT NURSING ASSISTANTS Hemmingway H, Stafford M, Stansfeld S, Shipley M, Marmot M. Is the SF-36 a valid measure of change in population health? Results from the Whitehall II study. BMJ 1997;315: Idler EL. Self-assessed health and mortality: a review of studies. 1992;1: Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. Journal of Health and Social Behaviour 1997;38: Labriola M, Lund T, Burr H. Prospective study of physical and psychosocial risk factors for sickness absence. Occup Med (Lond) 2006;56: Falkenberg A, Nyfjäll M, Bildt C, Vingård E. Predicting sickness absence are extended health check-ups of any value? Comparison of three individual risk models. J Occup Environ Med 2009;51: Voss M, Floderus B. How do job characteristics, family situation, domestic work and lifestyle factors relate to sickness absence? J Occup Environ Med 2004;46: Munch-Hansen T, Wieclaw J, Agerbo E, Westergaard- Nielsen N, Rosenkilde M, Bonde JP. Sickness absence and workplace levels of satisfaction with psychosocial work conditions at public service workplaces. AM J Ind Med 2009;52: