To evaluate the effectiveness of intervention programmes for the prevention of aggravation of back disorder or prolonged duration of sickness absence.

Searches were conducted of the following sources for studies published in the English language: MEDLINE from 1966 to January 1999; PsycLIT from 1887 to September 1998; EMBASE from 1988 to November 1998; CINAHL from 1982 to March 1998; Current Contents/Clinical Medicine from February 1998 to February 1999; Current Contents/Life Sciences from February 1998 to March 1999; CISDOC from 1987 to January 1999; HSELINE from 1987 to January 1999; MHIDAS from 1985 to January 1999; and NIOSHTIC from 1990 to January 1999. The search combined the keywords 'back-pain' or 'back disorder' with two other groups of keywords (groups 1 and 2): group 1 emphasised disability to work ('sickness absence', 'absenteeism', 'return to work', 'sick-leave', 'workers' compensation', and 'employment status') and group 2 focused on different types of intervention ('intervention', 'prevention', 'workplace', 'ergonomic intervention', 'design', 'vocational', 'education', 'lifting techniques', 'control', 'lumbar support', 'back-school', 'training', 'modified work', 'graded activity', and 'functional restoration'). In the search the keywords had to be present in combination with one or more words from the first group as well as from the second group. Relevant references in retrieved articles and published reviews were screened.

Study designs of evaluations included in the review

Studies had to contain a reference group. There was no selection on study design. Included studies were of randomised controlled design or prospective cohort design. Duration of follow-up ranged from 21 days to 5 years.

Specific interventions included in the review

Secondary types of non-medical prevention were compared to a reference treatment. Individual case management was excluded. Included interventions appear to be classified into two groups: back-school type interventions (a combination of exercise and functional conditioning, education, training in work methods and lifting techniques) and non back school type interventions (including McKenzie method exercise, functional conditioning and education, physician notification of treatment guidelines, combination of pain relief, and graded medical exercise). Reference treatments included: physical therapy/manipulation; placebo therapy; other care by physician; usual care; education only; conventional physiotherapy; self-exercise; and usual care and operative treatment. Organisational and administrative interventions were also searched for but none were found.

Participants included in the review

Patients with acute (pain lasting < 30 days), subacute (pain lasted between 30 days and 12 weeks) or chronic (pain lasted > 12 weeks) non-specific back pain or back disorders were eligible provided the duration of sick leave before the intervention did not exceed 1 year. Pregnant women with back disorders, patients with substantial co-morbidity with other diseases, people who were unemployed or on sick-leave of more than one year at the start of the investigation were excluded. Reports that did not concern assessment of disorders of the lower part of the spine known as the lower back were also excluded. Participants included factory workers, those referred to a medical centre or spine clinic, workers who had been compensated by insurance companies or compensation boards, those reporting work-related injuries and patients who were out of work.

Outcomes assessed in the review

The primary outcome was return to work (RTW). The impact of the intervention was also assessed by compliance, compliance sustainability and effect sustainability. Compliance was defined as the way in which subjects acted in accordance with the imposed intervention and was rated good if > 80% of participants complied, reasonable if between 30% and 80% complied, and poor if less than 30% complied. Compliance sustainability was defined as compliance with the intervention during the follow-up period and was rated good if > 80% of participants complied, reasonable if between 30% and 80% complied, and poor if less than 30% complied. Effect sustainability was defined as the way in which RTW was sustained without recurrence of new episodes of sickness absence during the follow-up period and was rated good if the percentage recurrence was 30% or less, reasonable for recurrence between 30% and 80%, and poor if recurrence was > 80%.

How were decisions on the relevance of primary studies made?

The authors do not state how the papers were selected for the review, or how many of the reviewers performed the selection.

The authors do not state that they assessed validity.

The authors do not state how the data were extracted for the review, or how many of the reviewers performed the data extraction.

Tables reported in the review included the following information: author and year of publication; study design; population; intervention; and results. The rate difference (RD) was calculated as the absolute difference between the percentage of RTW of the participants and the RTW of the referents. The preventable fraction (PF) was calculated as the relative difference between sickness absence of the referents and sickness absence of the participants divided by sickness absence of the referents.

How were the studies combined?

The studies were combined in a narrative review.

How were differences between studies investigated?

Differences between the studies were discussed.

Twelve studies were included (3873 participants), including nine RCTs (2269 participants) and three prospective cohorts (1604 participants).

The studies differed in design and population studied. Participants were heterogeneous and differed in duration and stage of low back pain at the start of the study. There were large differences among referent groups in inclusion and exclusion criteria and outcomes.

Compliance with the intervention was good in eight studies, reasonable in two studies, poor in one and unknown in another.

Compliance sustainability was rated good in three studies, poor in one study and unknown in the others.

Effect sustainability was rated good in three studies, reasonable in three studies, and unknown in four.

Absolute reduction of sickness absence and time lost from work ranged from 22% to 42%. The preventable fraction ranged from 50% to 70%.

Back-school type interventions (8 studies).

Studies differed in the period of follow-up (range 21 days to 1800 days).

Seven of the eight studies showed a significant overall difference between RTW of the participants and the control group. One study showed no statistically significant difference between groups.

Outcome < 60 days (4 studies): rate difference ranged from -7% to 29%. Preventable fraction ranged from -11% to 47%.

Outcome from 60 days to 1 year (8 studies): significant effect found in 4 studies, with rate difference 22% to 42%.

Outcome at 360 days (4 studies): rate difference ranged from -1% to 42%. Preventable fraction ranged from -100% to 76%.

Non back-school type interventions (4 studies).

Results were inconsistent, with only one study reporting a significant effect of the intervention. In three studies, the RD was better among referents than in the intervention group. Problems with studies included: poor compliance of the physician with the intervention (applying treatment guidelines) with few patients receiving the intervention; and intervention was not work related, was started too soon after sick leave due to back pain, and meant the end of receiving benefits.

Few studies were performed to assess the outcomes return to work after ergonomic intervention. However, there is evidence that intervention in the subacute phase of back pain is preferable. Future intervention studies should address intervention sustainability and recurrence of sickness absence due to back pain over at least 1-year follow-up period.

The aims were stated and inclusion criteria defined in terms of intervention, participants and outcome. A number of relevant databases were searched, although by restricting the search to articles published in the English language, other relevant studies may have been omitted. No attempts were made to locate unpublished material and methods used to select studies were not described. Study design was not specified in the inclusion criteria and validity was not assessed. Some relevant details of the primary studies were presented in tabular format but methods used to extract data were not described. Given the clinical heterogeneity among studies, a narrative review was appropriate. However, attention was not drawn to the better sources of evidence and statistical heterogeneity was not assessed.

In view of the lack of assessment of validity, the authors conclusions must be interpreted with caution.

Practice: The authors state that back school interventions, regardless of their programmes and heterogeneity, showed more effect after 60 days of sickness absence than other non back-school interventions, and that interventions in the subacute phase seemed preferable unless a strong intervention effect could be exercised upon the already strong recovery among patients in the early stages of low back pain.

Research: The authors state that future research should be aimed at RCTs concerning an organisational and technical intervention sustained over a follow-up period of at least one year, with special attention to the occurrence of low back pain sick leave.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.