1.1

Introduction

Return to work of low back pain (LBP) patients remains a public health issue as recently stated by the federal centre for healthcare expertise, KCE .

Chronic LBP is a common condition: a survey conducted in the area “Pays-de-la-Loire” showed that 58% of workers had suffered from LBP in the past year, and that 7% reported that pain was present every day. A recent report by Leclerc et al. highlights the social consequences of LBP . The main issue is the occurrence of major disability in a limited number of patients and this was already pointed by Spitzer et al. more than 20 years ago .

Several different approaches have been developed during the past two decades. The risk factors for chronicity have been extensively studied and therapeutic strategies have been developed. Psychosocial characteristics, among which the fears and believes concerning movement, have been shown to play a major role. Counselling strategies have been increasingly focused on these elements. LBP programs have been developed, based on the concept of physical and psychosocial deconditionning. They include multidisciplinary interventions with the objective of resuming activities and returning to work. Short-term outcomes such as improvement of physical performance are very significant . Evaluation of such programs includes the cost–effectiveness evaluation and requires long-term follow-up and data on the long-term work status. Such data are quite scarce and most of the available studies are limited to one-year follow-up .

Since 2000, several cohorts of LBP patients have been studied . Results at 1 year of a series of 87 patients showed a 60% decrease of sick leaves and 72% of the patients had returned to work . The objective of the present study is to report the results at 1 years and define the main factors predictive of return to work.

1.2

Patients and methods

This is a prospective open study. The inclusion criteria were:

• •
  

  age between 18 and 53;

  
  
• •
  

  presence of chronic LBP or lumbo-radiculalgy according to the criteria of the French health evaluation agency ;

  
  
• •
  

  presently employed with a permanent contract;

  
  
• •
  

  on sick leave or having been on sick leave for over 6 months in the past 2 years.

Exclusion criteria were secondary LBP, osteoarthritis or neurological disease precluding physical exercise, cardiovascular disease (diagnosed after cycloergometer stress tests) and psychiatric disorders incompatible with the participation in a group program.

1.3

Intervention

The intervention includes participation in a functional restoration program (FRP) and ergonomic intervention on the workplace. The FRP program was standardised and performed in two rehabilitation centres. It has been described in detail in the study reporting results at 1 year . This program includes physiotherapy interventions, occupational therapy interventions and psychological counselling. The medical supervisor is a physiatrist. Patients participate as in- or outpatients 6 hours per day, 5 days a week during 5 weeks.

The ergonomic intervention is provided on the workplace by a regional agency devoted to the improvement of work conditions (ARACT), according to the standards defined for “short interventions” by the national agency (ANACT). This intervention is only possible after agreement of the company and was performed for 53 patients. It consists in:

• •
  

  a diagnosis of the work conditions, conducted with the different actors;

  
  
• •
  

  the definition of an action program;

  
  
• •
  

  feedback for the company.

This intervention is tailored to the specific context of the company and cannot be completely standardised.

1.4

Evaluation

The evaluation is similar to that conducted at 1 year . Finger to floor distance was used to evaluate flexibility. The strength of the trunk flexors and extensors was evaluated by the Ito et al. and Sorensen tests. Lifting ability was evaluated by the “Progressive Isoinertial Lifting Evaluation” (PILE) test . Quality of Life was evaluated by the self-administered Dallas questionnaire . Disability was evaluated by the Quebec self-administered questionnaire . The Hospital Anxiety and Depression scale (HAD) was used to evaluate psychological consequences of LBP.

Sick leaves due to LBP were self-reported by the patients.

1.5

Statistical analysis

Mean value and standard deviation were calculated for quantitative variables. The chi ² test was used to compare qualitative variables and the T test for quantitative variables. Non-parametric tests for paired variables (chi ² and Wilcoxon test) were used for to compare the pre- and post-treatment values of qualitative variables. Pre- and post-treatment c quantitative variables were compared by the student t test.

Analysis of the predictive factors of return to work was performed by univariate analysis for the 61 patients in whom the work status at 2 years follow-up was available. We tested the following variables: gender, age, work constraints, previous spinal surgery, being on sick leave at the beginning of the program, having been on sick leave for over 1 year during the past 2 years, LBP due to industrial accident, Dallas score, HAD score, VAS pain score, physical criteria (FFT, Ito and Sorensen scores), self-assessment of improved physical condition at the end of the program, self-assessment of ability to return to work; self-assessment of ability to resume leisure and sports activities. Multivariate analysis was then performed using logistic regression modelling .

Quantitative variables were dichotomised as above or below the median value in order to obtain balanced classes. The following variables: gender, sick leave, implementation of the ergonomic intervention were forced into the logistic regression model. All variables which were related to the work status in the univariate anaysis with a correlation above 20% were included in the multivariate analysis. Statistical significance was set at 5%.

Analysis were performed on the SPSS software for Windows version 13.0.

1.6

Results

At 1 year, 83 patients completed the evaluation; one had not completed the program; two refused to come back for evaluation and one was lost to follow-up ( Fig. 1 ).

Chart flow diagram.

At 2 years, 81 patients (94%) completed the evaluation (one patient refused and one was lost to follow-up).

Twenty were excluded from further analysis for the following reasons:

• •
  

  11 had stopped working for medical reasons unrelated to LBP;

  
  
• •
  

  four had undergone surgery (one spinal cord injury, one intestinal occlusion, two knee sprains);

  
  
• •
  

  four were hospitalized for psychiatric reasons;

  
  
• •
  

  three were off work due to other medical conditions;

  
  
• •
  

  nine had stopped working for economic reasons :

  
  
  
  
  • ∘
    

    one female patient stopped work after her husband had a promotion,

    
    
  • ∘
    

    three patients were fired for economic reasons,

    
    
  • ∘
    

    five patients followed occupational training programs).

  
  

The analysis of the factors which were predictive of the work status was conducted in the 61 remaining patients (70%).

There was no significant difference in gender, age, duration of sick leave before treatment, nor in any of the outcome measures between the 21 patients excluded and the 61 patients included in the final analysis.

Forty-four patients over 61 were males; the mean age was 40.8 years. Eighty-four percent were employed in economic sectors characterized by heavy work constraints. The mean duration of sick leaves in the 2 years prior to the program was 329 ± 179 days. Seventy-two percent were on sick leave when included in the program.

Ergonomic intervention was conducted in 41/61 (67%).

At 2 years, 48 (79%) completed all tests (physical items, self-questionnaires, data on work status). The 13 others completed a telephone interview and gave information concerning their work status. Seven of them also completed the self-assessment of physical condition, of the ability to work and of the ability to resume sports activities.

Tables 1 and 2 summarize the physical data.

All outcome measures remain significantly better at 2 years follow-up than before treatment ( p < 0.01) ( Table 1 ).

Fifty-six over 61 (92%) “feel able to work” at the end of the program; 53/57 (93%) at 1 year and 46/55 (84%) at 2 years.

Sixty-one over 61 (100%) « feel their physical condition is improved » at the end of the program, 44/57 (77%) at 1 year and 37/55 (67%) at 2 years.

Fifty-three over 61 (87%) “feel able to resume sports activities” at the end of the program, 38/57 (67%) at 1 year and 25/55 (45%) at 2 years.

Less than half of the patients still had a physical activity 2 years after the program. Despite this fact, two thirds felt that their physical condition was better and three out of four felt able to work. Fig. 2 shows the decrease of the items “improved physical condition” and “physical activity” at 1 year whilst 80% still “feel able to work” at 2 years.

Self-assessment.

1.6.3

Evolution of sick leaves

The mean duration of sick leaves decreases by 60% in the 2 years following the program, compared to before the program (128 ± 200 days versus 329 days ± 179 days).

Fourteen patients (23%) have been on sick leave at least 6 months in the two following years. The number of sick leaves is not significantly different during the second post-program year, compared to the first ( Fig. 3 ) ( Tables 4 and 5 ).

Duration of sick leaves.

Table 4

Factors significantly correlated to the work status at 1 and 2 years.

Variables	At work at one year		p	At work at two years		p
	( n = 62)	%		n = 48	%
“feels able to work” T5
Yes	60	81.1	< 0.001	46	82.1	< 0.06
No	2	22.2		2	40.0
Ergonomic intervention
Yes	45	86.5	< 0.002	33	80.5	< 0.6
No	17	54.8		15	75.0
Dallas T1 > median value
Yes	26	60.4	< 0.002	24	77.4	< 0.8
No	36	90.0		24	80.0
Anxiety/dépression Dallas T5 > median value
Yes	27	64.3	< 0.003	28	73.7	< 0.3
No	35	85.4		19	86.4
Anxiety/ Depression Dallas T1 > median value
Yes	26	61.9	< 0.007	22	68.8	< 0.07
No	35	87.5		25	89.3
ADL Dallas T1 > median value
Yes	28	63.4	< 0.013	24	77.4	< 0.8
No	34	87.2		24	80.0
Industrial accident
Yes	23	92.0	< 0.014	14	77.8	< 0.9
No	39	67.2		34	79.7
Dallas T5 > median value
Yes	27	64.2	< 0.024	(25)	78.1	< 0.9
No	35	85.4		(23)	79.3
Sick leave before treatment ≥ 6 months/1 year
Yes	19	65.5	< 0.037	17	81.0	< 0.5
No	42	85.7		31	77.5
Work/Leisure Dallas T5 > median value
Yes	29	84.6	< 0.043	26	76.5	< 0.6
No	33	65.9		22	81.5
“has resumed sports activities” T5
Yes	29	84.6	< 0.043	44	83.0	< 0.06
No	33	65.9		4	50.0
HAD depression T1 > median value
Yes	24	70.6	< 0.07	24	72.7	< 0.07
No	27	90.0		23	92.0
PILE %body weight T1 > median value
Yes	33	82.5	< 0.09	28	90.3	< 0.03
No	29	67.4		20	66.7
Work/leisure Dallas T1 > median value
Yes	32	69.6	< 0.19	24	75.0	< 0.5
No	29	80.5		23	82.1
HAD depression T5 > median value
Yes	25	73.5	< 0.4	24	72.7	< 0.2
No	28	84.8		24	85.7

 

Table 5

Factor significantly correlated to the work status at two years.

Variables	OR	IC 95%	p
PILE %bw T1 ≥ 50%	8.2	1.5 45.3	0.015
HADd T1 ≥ 50%	0.2	0.0 1	0.052

 

Logistic regression. The variable included in the model were: gender, ergonomic intervention, Sick leave less than 6 months in the two previous years, « feels able to work at T5 », « feels able to resume sports activities at T5», A/D Dallas T1 ≥ 50%, A/D Dallas T5 ≥ 50%, PILE % bw T1 ≥ 50%, HADd T1 ≥ 50%, HADd T5 ≥ 50%.

OR: odd ratio; IC _95% : 95% confidence interval of the odd ratio

We analyzed the factors associated with the work status at 2 years and compared them to what had been found at 1 year. This analysis was performed in the 61 patients in which all data were available. Forty-eight of them were at work. The factors were compared to those identified at 1 year (83 patients, 62 of them were at work).

PILE above the median value at the beginning of the program (PILE T1) is the only factor significantly ( p < 0.03) correlated with the work status at 1 years when univariate analysis is performed. Multivariate analysis reveals a significant positive correlation of PILE T1 and a significant negative correlation of the depression subscore of the HAD scale.

1.7

Discussion

This study shows that over 78% of the patients included in the program remain at work at 1 years. The proportion was of 72% at 1 year. This is in agreement with the results of the meta-analyses performed by Hayden et al. , by Kool et al. , and with the studies by Friedrich et al. and Verfaille et al. . In agreement with these previous data, our results show the efficiency of such programs, on the medium- and long-term. Despite overall satisfactory results, 13% of the patients remain on sick leave for LBP or lumbar surgery. When patients on disability pension are included in the count, the proportion of patients off work for over 6 months reaches 23%. Proctor et al. also report that a small proportion of patients remain in unsatisfactory situations despite selection criteria at the entry of the program.

The main methodological difficulty is the analysis of the status of patients who do not work but in which the reason for the cessation of activity is not LBP. We chose to exclude them from the final analysis . We also chose to exclude the five patients who where re-training because the delay and duration of this retraining was very variable. The sub-group of patients excluded from the final analysis was comparable to the general population for all criteria. This procedure reduces the power of the study and possibly over-estimates the decrease in the number of sick leaves.

The only factor correlated with the number of sick leaves is the performance on the lifting task before the program. This has not, to our knowledge, been previously reported and deserves confirmation. Factors which have been previously reported to be correlated such as: “feels able to work”, “has resumed physical activities”, or Dallas or HAD scores above the median value do not, in our study, reach significance. The increase in the physical or psychological criteria during the program does not appear to be correlated with the work status at 2 years either. Ergonomic intervention is not significantly correlated to the work stats at 2 years and we found no clear relation between ergonomic intervention and effective changes in the work environment. Although ergonomic intervention has been described as one of the tools to promote return to work, this does not appear in this study . This may be in part due to the difficulties in standardising this intervention. The possibility of such an intervention heavily depends on work organisation characteristics which probably in themselves influence return to work and this is a major methodological issue.

After decreasing in the 6 months following the program, Physical criteria remain stable and remain better at 2 years than before the program. This is achieved despite the fact that more than half of the patients do not perform physical activities. The absence of correlation between pursued physical activity and the presence at work has already been described by other authors .

The determination of risk factors for persisting LBP and persisting work disability is extremely difficult even when meta-analyses are performed . This discrepancy is in part due to the role of other factors. Kool et al. stress the importance of coping strategies developed during such programs. Recently, Mc Geary et al. highlight the role of pain intensity and reports that high pain intensities before the program are often associated with bad outcomes both in LBP and other musculoskeletal diseases. The debate on the respective influences of depression and anxiety and of fear and believes remains open .

For Wessels disabilities play a greater part in return to work than impairments; but Chambon et al. and Smeets et al. report a decrease of <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='VO2max’>VO2maxVO2max
V O 2 max
values in chronic LBP.

Sullivan et al. highlight the importance of the work environment in addition to the psychosocial context in determining the probability to return to work. This could be a major issue and lead to a double approach combining individual components but also the reality of the work context as proposed in the predictive factors developed by Proctor et al. . Over-reaction creates disability more surely than purely physical impairments as described by Genêt et al. . This environmental approach of the LBP worker has been particularly developed by Loisel and justifies the concept of “therapeutic return to work” . In a similar approach, Linton et al. include the family and relatives of the patients in the context .

In chronic LBP, networking is necessary and should include rehabilitation facilities, occupational specialists, and the work environment. This allows communication of the conclusions of the program and progressive return to work and requires that all information can be shared between the patient, relatives, service providers, healthcare facilities as in some of the Canadian models .

1.8

Conclusion

Despite the methodological difficulties, these results show that the benefit of a functional restoration program remains significant at 2 years follow-up in terms or reduction of sick leaves.

Determining factors correlated with the return to work is difficult, in part because of the lack of power of our study but also because many different factors coexist, are interdependent, none of them being in itself determinant.

2.1

Introduction

Le problème du retour à l’emploi des lombalgiques chroniques demeure un problème de santé publique de premier ordre comme en témoigne le récent rapport du Centre fédéral d’expertise des soins de santé belge, KCE .

La lombalgie chronique est fréquente : 58 % des salariés de Pays-de-la-Loire disent avoir souffert de lombalgies dans l’année et 7 % de façon quotidienne d’après une étude épidémiologique du réseau de surveillance des TMS . Ses conséquences sociales sont mises en lumière par l’équipe d’Annette Leclerc dans un récent rapport de l’Inserm . De plus, la caractéristique essentielle de ce véritable fléau est la survenue d’incapacités prolongées dans un nombre réduit de cas comme l’illustrait déjà le rapport Spitzer et al., il y a 20 ans .

Au cours des deux dernières décennies, plusieurs types d’approche ont été développés. De nombreux travaux ont analysé les facteurs de risque de passage à la chronicité et proposé des stratégies thérapeutiques. Le caractère central des éléments psychosociaux, notamment des peurs et croyances relatives au mouvement, a été mis en évidence. Les stratégies d’information thérapeutique ont été de façon croissante ciblée sur ces aspects . Des programmes de traitement des patients lombalgiques chroniques ont été développés . La majorité reposent sur le concept de déconditionnement physique et psychosocial et proposent des interventions multidisciplinaires dont l’objectif est la reprise ou le maintien dans l’activité, notamment professionnelle. Les résultats immédiats de ces programmes sur les limitations des performances physiques initiales sont très significatifs . L’évaluation de ces programmes, notamment dans son versant médicoéconomique, nécessite aussi un suivi, permettant d’analyser le parcours, en particulier professionnel au long cours. Ces données restent actuellement relativement peu disponibles ; la plupart de ces études disposent d’un an de recul et rares sont celles de plus d’un an .

Plusieurs cohortes de patients ont été suivies dans la région de Pays-de-la-Loire depuis 2000 . Les résultats à un an d’une cohorte de 87 patients montraient une diminution de 60 % de la durée des arrêts de travail et un maintien dans l’emploi pour 72 %. L’objectif de la présente étude est d’analyser l’évolution de ces résultats au cours de la deuxième année et de tenter de dégager des facteurs prédictifs de retour ou de maintien dans l’emploi de cette population à haut risque d’exclusion.

2.2

Patients et méthodes

2.3

Résultats

À un an, l’évaluation a porté sur 83 cas, car un cas n’a pas achevé le programme, deux ont refusé d’être suivis, un a été perdu réellement de vue ( Fig. 1 ).

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Dispositif LMD en médecine et métiers de la rééducation : le système européen s’applique en France Interexaminer agreement of clinical examination of the neck in manual medicine Effects of a primary rehabilitation programme on arterial vascular adaptations in an individual with paraplegia Sudden dysphagia in an elderly, quadriparetic patient Multicenter multidisciplinary training program for chronic low back pain: French experience of the Renodos back pain network (Réseau Nord-Pas-de-Calais du DOS) The effect of music therapy on mood and anxiety–depression: An observational study in institutionalised patients with traumatic brain injury

Stay updated, free articles. Join our Telegram channel

Join