«FICHA TÉCNICA Título Segurança e Higiene Ocupacionais - SHO 2012 - Livro de Resumos Autores/Editores Arezes, P., Baptista, J.S., Barroso, M.P., ...»
After that, we investigate what factors related to the home environment, were associated with complaints in the lumbar region. From the variables selected using the univariate binary logistic regression models, were introduced in the prediction model those who could contribute to the dependent variable (complaint in the lumbar region). We used stepwise forward selection method. This process resulted in eight variables that have a contribution to the risk of having complaints in the lumbar region. The contribution may be positive or negative depending on the sign of the estimated
coefficients. The forecast model can be seen in Equation (1):
logit = 2.719 + 1.584 *X1 – 2.222*X2 – 1.237*X3 + 2.093*X4 + 1.187*X5 - 3.404*X6 – 4.047*X7 – 2.056*X8 (1) X1-complaints in the hands / wrists; X2-complaints in thighs; X3-posture of the forearm; X4-repetitive movements; X5posture of the arm; X6- the arm, or its weight is supported; X7- devices aids for lifting/transferring patients; X8-job satisfaction.
The model performance was evaluated by ROC analysis yielding a value for the area under the ROC curve of 0.889 (p0.05), which reveals a high discriminating power, that is, the model is able to correctly predict the complaints in the lumbar region in 88.9% of cases.
4. CONCLUSIONS Given that the number of responses received so far does not allow an inference about the population, however we can
characterize this sample according to some important aspects, namely:
The body zones with the greater number of complaints are the back (cervical – 73.5%, lumbar – 64.6%, dorsal – 49.0%) and shoulders (49.0%). These values are somewhat consistent with studies of other authors carried out both at hospital context and at homecare settings (Cheung et al. 2006; Smith et al. 2006).
About 80.8% of home care nurses consider the height of the bed (or any other surface where’s the patient) low;
18.4% consider it suitable and 0.8% consider it high. This is an important aspect because previous studies already revealed as a factor in the emergence of awkward postures and consequently of musculoskeletal complaints (de Looze et al. 1994).
We find a statistically significant association between “musculoskeletal complaints in the lumbar region” and “provide home-based care”, (OR=3.19 (p0.05), 95% Confidence Interval [1.26; 8.08]). So, the nurses who provide home-based care have circa triple chance of having musculoskeletal complaints in the lumbar region than their colleagues.
The statistical model obtained for risk forecasting is one that has the greatest discriminating power in terms of area under the ROC curve. The value of the area under the curve is 0.889, meaning that the model is able to correctly predict the complaints in the lumbar region in 88.9% of cases, to the nurses who provide home-based care.
5. REFERENCES Barroso, M., Carneiro, P., Braga, A.C. (2007). Characterization of Ergonomic Issues and Musculoskeletal complaints in a Portuguese District Hospital. In Proceedings of the International Symposium “Risks for Health Care Workers: prevention challenges”. Athens.
Björkstén, M.G., Boquist, B., Talbäck, M., Edling, C. (1999). The validity of reported musculoskeletal problems. A study of questionnaire answers in relation to diagnosed disorders and perception of pain. Applied Ergonomics, 30, 325-330.
Cheung, K., Gillen, M., Faucett, J., Krause, N. (2006). The Prevalence of and Risk Factors for Back Pain Among Home Care Nursing Personnel in Hong Kong. American Journal of Industrial Medicine, 49:1, 14-22.
de Looze, M.P., Zinzen, E., Caboor, D., Heyblom, P., van Bree, E., van Roy, P., Toussaint H.M., Clarijs, J.P. (1994). Effect of individually chosen bed-height adjustments on the low-back stress of nurses. Scandinavian Journal of Work, Environment & Health, 20, 427-434.
Hignett, S. and McAtamney, L. (2000). Rapid Entire Body Assessment (REBA). Applied Ergonomics, 31:2, 201-205.
Kuorinka, I., Jonsson B. and Kilborn, A. (1987). Standardized Nordic Questionnaires for Analysis of Musculoskeletal Symptoms.
Applied Ergonomics, 18:3, 233-237.
Smith, D.R., Mihashi, M., Adachi, Y., Koga, H., Ishitake, T. (2006). A detailed analysis of musculoskeletal disorder risk factors among Japanese nurses. Journal of Safety Research, 37, 195-200.
Occupational Safety and HygieneInternational Symposium on
Safety at work and worker profile: analysis of the manufacturing sector in Andalusia in 2008 Carrillo, Jesusa, Gómez, María Almudenab, Onieva, Luisc a Junta de Andalucía, Avenida Hytasa 14, 41006 Sevilla email: email@example.com ; aInstituto Andaluz de Prevención de Riesgos Laborales, Avenida Hytasa 12, 41006 Sevilla email: firstname.lastname@example.org; cUniversidad de Sevilla, Camino de los Descubrimientos s/n, Sevilla e-mail: email@example.com
1. INTRODUCTION According to the Encyclopaedia of International Safety Organization, Part VIII, Chapter 56, causes of accidents should be classified as immediate causes (unsafe acts, unsafe conditions) and contributing causes (safety management performance, mental condition of workers and physical conditions of workers). This model implies that individual worker characteristics can affect occupational safety in terms of proneness to injuries because they are related to physical and mental conditions. Within the same company, activity, task and job, there would be personal risk factors that can affect the likelihood of accidents. The identification of those factors can help to develop better preventive programs and optimize resources. Public prevention programs can be designed and adjusted to those collectives of workers at risk.
Future Public Strategies should propose specific actions for them.
In this paper we analyse which worker characteristics can constitute risk factors in the Andalusian Manufacturing Sector (NACE 15 to 27, Council Regulation EEC Nº3037/90) based on a cross-sectional survey of working Conditions (Dembe, Erickson, & Delbos, 2004). Our purpose is to determine if there are worker characteristics that influence accidents in Manufacturing Sector in Andalusia like age or gender, controlling possible confounders such as differential employment of each worker group or company characteristics.
There is no previous information in Andalusia about which worker characteristics could influence injury rates in the manufacturing sector. Explanatory variables have been selected based on a literature review. Main variables should be worker occupation and the company’s main activity because they determine most of the injury hazards at workplace related to immediate causes (unsafe conditions). Other possible explanatory variables at company level are company size (Sorensen, Hasle, & Bach, 2007), work shifts (Folkard & Tucker, 2003) and type of contract (Benavides, Benach, Muntaner, Delclos, Catot, & Amable, 2006).
Individual worker risk factors that should be taken into account are age, gender, nationality and seniority, according to previous studies. The majority of studies have reported that young workers had a higher injury rate, especially if they are men. However, injuries among young workers seem to be less severe (Salminen, 2004). There are important differences in terms of job assignment for each country and activity and foreign workers are expected to be employed in more dangerous tasks (Ahonen, Benavides, & Benach, 2007). Female workers show lower injury rates as a general trend (Islam, Velilla, Doyle, & Ducatman, 2001), however, some authors propose that this can be explained because of differential jobs and task assignment (Smith & Mustard, 2004). In fact, when analyzing at company level, female workers have more accidents than their male counterparts (Kelsh & Sahl, 1996) even with the same jobs and tasks.
As a first insight and reference, we have calculated raw injury rates based on official notifications of accidents and in the estimation of number of workers gathered by National Employment Survey (“Encuesta Nacional de Población Activa”) performed by National Institute of Statistics (“Instituto Nacional de Estadística”) and available at http://www.ine.es.
Accident notifications are electronically collected through Delt@ information system in Spain for all accidents that result in an absence from work of one or more days. Stratified information in terms of age, sex and tenure for Andalusia is available only for some manufacturing activities (NACE 15 to 22 and 26 to 37). Raw injury rates are higher for young workers, inexperienced and male workers (see Table 1).
2. MATERIALS AND METHODSWe have used data from First Andalusia Working Conditions Survey (Instituto Andaluz de Prevención de Riesgos Laborales, 2009). Respondents were interviewed at home with a questionnaire of 81 items. Sampling of respondents was adjusted for province, activity and sex, with a global error estimated in ±1.01%. Survey micro-data and regression models are available upon request for other researchers. Only those cases who declare being employed in the manufacturing sector have been analyzed (909 cases). For each case, there are available several explanatory variables, based on their reported answers.
Workers were asked if they had been involved in at least one accident during the last two years. This variable is used as the dependent variable in logistic regression analysis of data (108 cases reported accidents). We have calculated the odds ratio, both univariate and multivariate. For each variable we have classified cases according to categorical values, one of these categorical values is used as reference for odds ratio calculation (therefore OR is 1.00 for reference variables). Most authors consider that, as a rule of thumb, 10 cases are needed for each included variable in the model (9 variables).
3.1. Univariate analysis of First Andalusian Working Conditions Survey.
The categorical variables were checked as explanatory risk factors (Table 2). According to the univariate analysis, risk factors are: male workers, workers over 45 years old, workers in big companies, technicians, workers in companies with worker representation, workers with more than 45 hours per week, workers with night or rotating shifts, workers of subcontractors, workers without an university degree and workers who reported that their main risk was falling, mechanical or working place. In univariate analysis we do not find any statistical significance for nationality, type of company (private or public) or training hours in the last two years.
In univariate analysis, some variables show no significant relation with accidents. Contract type, outsourcing and workers with hygiene as main risk have no significant relationship. Explanatory variables to be included in a multivariate regression are sex, age, company size, worker representation, education level, occupation, tenure, main risk, weekly working hours and shift type.
3.2. Logistic regression analysis of First Andalusian Working Conditions Survey.
As the survey is a cross-sectional study, multivariate logistic regression can be performed to check if there are possible confounders of the univariate relationships. We tried several options; finally two models were adjusted (with SPSS).
After adjusting the model, see results in Table 3, personal characteristics such as sex, age and education are no longer significant. Also, no evidence could be found to support the idea that company size, worker representation, shift type or subcontracting are safety risk factors. Using the adjusted model, risk factors are occupation as technician, tenure longer than 6 months, working over 40 hours per week and identification of mechanical as their main risks.