2 Considerations

The PDG took account of a number of factors and issues when developing the recommendations.

Definitions and terms

2.1 The evidence reviews that inform this guidance identified any relevant interventions, policies, strategies or programmes to help people return to work after sickness absence and/or incapacity. For the purposes of this guidance, the term 'intervention' has also been used to cover policies, strategies and programmes. 'Incapacity' has been used to mean long-term inability to work because of illness or disability.

2.2 The original DH referral asked NICE to develop public health guidance for managing long-term sickness and incapacity. However, there is no consensus in the literature on how to define long- or short-term sickness absence. For this guidance, short-term sickness absence has been defined as absences from work of up to (but less than) 4 weeks, and long-term sickness absence as lasting 4 or more weeks. The criteria for qualifying for incapacity benefit have changed over time. In October 2008, a new employment and support allowance (ESA) was introduced which will eventually replace incapacity benefit and income support on grounds of incapacity. To ensure potentially relevant studies were not missed, the terms short-term sickness absence, long-term sickness absence and sickness absence were all used in the literature search. Studies that included participants receiving incapacity benefit or a similar benefit were also included.

2.3 The PDG recognised that people who take significant cumulative absences (such as multiple short-term sickness absences linked to a specific condition) are probably more at risk of long-term sickness absence then those taking occasional single day absences. It also noted that the causes of short-term and long-term sickness absence are likely to differ; some conditions, for example back pain, are more likely to result in long-term sickness absence and acute medical conditions are more likely to result in short-term sickness absence.

2.4 The lack of studies clearly defined as covering 'long-term sickness absence', 'short-term sickness absence' or 'recurring short- or long-term sickness absence' has meant that the PDG has not always been able to produce recommendations that distinguish between these terms. In future studies, it will be important for researchers to define the terms they use and use them consistently. In particular, it is important to clarify the duration of long and short-term sickness absences, for both full and part-time employees.


2.5 The recent review of the health of Britain's working-age population was based on the premise that work has inherent benefits for people's health. It also recognised gaps in the evidence on how effective and cost-effective work-based interventions and health interventions are in promoting a return to work (Health, Work and Wellbeing Programme 2008). This guidance complements the proposals identified in the review and reiterates the importance of addressing the gaps in the evidence base for this topic.

2.6 An individual's health is the result of a set of complex interactions between multiple biological and social factors, including, for example, their:

  • sex, biological predisposition and genetic traits

  • socioeconomic position

  • access to information, services, support and resources

  • exposure to risk, including environmental risk factors

  • degree of control over their own life circumstances

  • access to (and their interaction with) the healthcare system (Marmot and Wilkinson 2005).

    All these factors affect people's ability to withstand the stressors – biological, social, environmental, psychological and economic – that can trigger ill health (Marmot and Wilkinson 2005). If an individual is absent from work for a prolonged period of time due to health reasons, then it is likely that more than one of these factors will have contributed to their absence. Furthermore, the number of people claiming incapacity benefit is greater in areas of higher unemployment, slower economic growth and higher socioeconomic deprivation (Beatty and Fothergill 2005; Norman and Bambra 2007). People receiving incapacity benefit are less likely to have academic or professional qualifications than those in work (McCormick 2000). As a result, they are likely to need education and training before they can achieve sustainable employment. This suggests that health is only one of the factors that will need to be addressed when helping someone return to work (Black 2008).

2.7 The PDG recognised that the workplace, including employer and employee practices, may contribute to or cause someone's absence from work due to sickness. (An example of such practices includes those which discriminate against certain groups or which do not adequately protect people's health and safety.) Consequently, both employers and employees have an important role in helping people get back to employment after long-term sickness absence and incapacity. This includes ensuring recruitment and selection practices do not exclude or discriminate against those who have experienced long-term sickness absence and incapacity. It may also include an assessment of the person's current fitness for employment and making workplace adjustments as required by health and safety legislation and the Disability Discrimination Act (HM Government 1995/2005). It may also include the provision of any re-training as needed. Campaigns and schemes such as 'Job introduction scheme', 'Job interview guarantee', 'Shift' and 'Mindful employer' aim to overcome stereotypes and stigma about disability, ill health and its effect on employment and employment opportunities. The PDG noted that specialist job advisers, such as Jobcentre Plus staff, may also be able to offer advice and support. It also noted that the Access to Work Scheme can help fund reasonable adjustments to the workplace for employees with disabilities.

2.8 Different types of employer (such as large, small or public and private organisations) are likely to have different policies and practices on sickness absence, which means the criteria and trigger points for intervening may differ. For example, the number of days of sickness absence before a sickness absence policy is triggered may vary. Consequently, employers implementing the recommendations may need to consider adjusting their employment contracts and/or organisational policies.

Inclusion and exclusion criteria

2.9 Study design was a key component of the inclusion criteria for the evidence reviews. A wide range of studies that assessed data measures before and after delivery of the intervention were included. They ranged from randomised controlled trials, before and after studies (with and without controls) and case–control, cohort studies to expert papers (see appendix B). These are the most appropriate study designs for determining causality between intervention and effect. Given the limited time frame, descriptive studies examining the relationship between ill health and sickness absence or incapacity were excluded. Similarly, qualitative studies (such as in-depth interviews and focus group data describing participants' views and experiences) were excluded. The PDG was aware, however, that such studies might provide data that would complement the effectiveness and cost-effectiveness data (for example, information on the barriers to and facilitators of delivery for specific interventions).

2.10 To ensure the literature searches and evidence reviews used to inform the guidance corresponded with the referral received (and to work within available resource) a number of exclusion criteria were applied (see appendix B). For example, the following were excluded:

  • research not published in English

  • dissertations, books and book chapters (however, the findings from such sources may also be available in journal publications)

  • interventions assessing the effectiveness of private health insurance schemes – although an intervention delivered by private health insurance companies would be included if it involved a workplace or primary care partner

  • interventions assessing 'ill-health retirement' and their outcomes

  • studies on fiscal policies, such as evaluations of disability working allowance and its impact on return-to-work outcomes

  • studies assessing the effectiveness of statutory or occupational pay schemes or studies assessing interventions which aim to prevent the first occurrence of sickness absence (primary prevention).

2.11 The evidence review covering interventions for people receiving incapacity or similar benefits was restricted to UK studies. There may be relevant international studies but differences in national policy, legislation and the benefits system mean it would not necessarily be feasible to implement the interventions in the UK.

2.12 The PDG identified a number of implications related to the inclusion and exclusion criteria:

  • Three of the evidence reviews used as the basis of the guidance were restricted to studies that only covered absences recorded as sickness and excluded studies on other absences (for example, maternity leave). Some studies that did cover sickness absence may have been excluded because of the lack of consistency in how employers record absences or lack of detail on the reasons for the absence.

  • All the evidence reviews were limited to interventions that involved employers and primary care providers (although they did not need to be the only providers involved, and the interventions could be delivered in various settings). A few studies were excluded because they did not describe explicitly who was involved in the intervention.

  • There may be studies covering population groups that were not specified in the inclusion and exclusion criteria or explicitly searched for, but which might have been useful to consider. Examples include: people who are employed but receiving incapacity benefit because they are no longer eligible for employers' sickness benefits; or interventions for those who are unemployed and receiving jobseekers allowance (or previous forms of this benefit) such as regional employability programmes.

  • Studies involving mixed population groups (such as self-employed and employed people or those experiencing sickness absence or other types of absence) would not have been included if it was not possible to disaggregate the data into a form that met the inclusion criteria. The PDG noted that the recommendations may also help self-employed and unemployed people return to work.

  • Similarly, mixed study designs (such as quantitative and qualitative) would not have been included for the same reason.

  • Interventions involving the clinical diagnosis, treatment and management of conditions that have resulted in sickness absence and/or incapacity were excluded as they were not part of the remit of this guidance. As a result, studies that also provided data on non-clinical interventions may have been excluded because the data presented were not sufficiently disaggregated.

  • Studies that did not report on return-to-work or work-related outcomes were excluded. A large number of studies were excluded from the evidence reviews for this reason.

  • Studies of return-to-work interventions that were planned, designed, delivered, managed or funded solely by local authorities were excluded. Similarly studies of return-to-work interventions that operate without any primary care or workplace involvement were excluded. (For example, this included some mental health-orientated strategies and studies on 'New deal for disabled people'.)

  • Although not part of the inclusion criteria for this guidance, it was noted that studies which examine the prevention of the first occurrence of sickness absence could prove valuable.

Collating and assessing evidence

2.13 A number of methodological issues were identified:

  • Some of the evidence considered originated from interim evaluations. When final evaluations of these activities are published, they may fill part of the gap in the evidence.

  • Work-related outcomes (rather than health) were the primary outcomes of interest for this guidance. However, improvements in work-related outcomes were not the primary outcomes or the main aim of some of the included studies (such as expert patient programmes). Consequently, data such as detailed statistics on return-to-work outcomes were often not reported.

  • A wide range of work-related outcomes would have been considered in the evidence review focusing on incapacity (such as return to paid and unpaid work, job seeking behaviour, increase in work experience or vocational training). However, the majority of the identified studies in the incapacity review only reported on outcomes associated with a return to paid work. Very few reported on other outcomes of interest.

  • Details were often not given about the content of the intervention, at what point during a person's sickness absence it was delivered, by whom, in what setting and how often and for how long. This made comparison across the different types of interventions difficult. It also made it difficult to identify exactly which elements of the intervention (for example, delivering it early in the absence) influenced its effectiveness.

  • Some studies lacked control groups.

  • Very few studies presented any cost or economic data.

  • Follow-up periods were variable (from weeks to months to years) and often details on the sustainability of interventions (1 year and beyond) were not reported.

  • Some studies involved multiple components and did not always report the differential effectiveness of each component.

2.14 A review commissioned by the Vocational Rehabilitation Task Group (Waddell et al. 2008) was published following the evidence consultation phase on this guidance. This review provides important evidence which will be considered (alongside any further published evidence) when this guidance is updated. (NICE guidance is usually updated every 3 to 5 years.)

Synthesising the evidence

2.15 Most of the evidence came from non-UK studies (in particular Scandinavian countries) and, where this is the case, the question of its applicability to England must be taken into account. Particular international variations include:

  • 'treatment as usual', which was used as the comparator in many studies

  • provision of financial incentives or compensation (for example, sickness benefits); these factors also vary between different types of employer

  • qualifications, roles and responsibilities or specific occupations (such as social workers) between countries

  • welfare benefits and their eligibility criteria

  • the descriptors used to report the reasons for sickness absence and incapacity; for example, the same condition might be categorised as linked to 'musculoskeletal disorders' in one country and 'stress-related' in another.

2.16 The PDG primarily relied on effect size and statistical significance to determine which interventions to include in the recommendations. However, in many studies the effect size and/or statistical significance at a 95% or 99% confidence interval was not reported. In such situations, if outcome data indicated general positive trends, the PDG considered making recommendations for practice.

2.17 Some of the studies reviewed indicated that intervening at an 'early' stage during sickness absence contributed to the success of the intervention. However, there is no universal definition of 'early' in terms of days, weeks or months. Consequently, where possible, the recommendations outline possible time periods to intervene. The Health and Safety Executive has produced a ready reckoner guide, which outlines the average lengths of absence by illness, by sector and by occupation, to help employers assess when to intervene. It may be useful for employers to refer to this guide alongside this guidance document.

2.18 Emerging evidence on the Expert Patients' Programme (expert paper 1) and condition management (expert paper 2) was considered. Both highlighted the need to help people overcome psychological or physical barriers before they can return to work. Condition management, when combined with informal employment advice, increased confidence levels about finding work and led to some increases in the number of people on incapacity benefit who returned to work. However, the PDG recognised that longer-term follow-up and evaluation was needed.

2.19 Relatively little evidence was identified on the effectiveness and cost effectiveness of interventions (such as those focusing on stress and mental illness or psychological interventions for specific population groups). Either they had not been evaluated or the evaluations were not publicly available. The PDG noted that the absence of evidence should not be taken as an indication that such interventions should be stopped (if they help to improve work-related or treatment outcomes). It also noted that some interventions may cause harm, even though there is no evidence to prove this. (See appendix D for further information on the evidence gaps.)

2.20 The experience, training and competencies of those coordinating or delivering the intervention/s – and their access to supervision and consultation with more skilled and higher qualified professionals – may affect the long-term effectiveness and cost effectiveness of any intervention.

2.21 Improving the quality of the evidence is a continuing process. Better evaluation processes are needed to help improve the available evidence base for this area (see section 5, research recommendations).

Cost effectiveness

2.22 Evidence on cost effectiveness was generally sparse. Most of the interventions dealt with musculoskeletal conditions and, in particular, lower back pain. Where evidence existed, it showed that such interventions were cost effective from both an NHS/personal social services perspective and a societal perspective. The analysis from the employer's perspective showed that, for the average employer, most of the effective interventions would, in the long run, reduce their costs. Usually this would be achieved through production increases attributable to earlier and/or a more effective return to work. These results can probably be applied to most employers (see appendix C for further details).

2.23 The cost effectiveness modelling for this guidance relates to interventions for those with a long-term sickness absence, excluding those in receipt of incapacity benefit. It was carried out at a time of relatively full employment. It assumed as a 'base case' that the average length of time it takes to replace someone who is on long-term sick leave was 10 weeks. In October 2008, a sharp increase in unemployment was forecast and the PDG recognised that, as a result, it may take less time to replace someone temporarily. Furthermore, when there is a larger pool of unemployed people there may be less of an imperative to get people back to work. As a result, interventions aimed at getting people back to work earlier will probably not be as cost effective as indicated in the modelling, which assumed conditions of low unemployment.

2.24 The PDG recognised that in times of severe or very severe economic downturn, these interventions will become less cost effective (that is, there will be a higher cost per QALY gained) than when there is full employment. However, it was not possible to determine whether they would be still be sufficiently cost effective to be a good use of government funds in times of relatively high unemployment. It also recognised that the probability of a return to work for those receiving incapacity benefits would decrease when there is high unemployment.

  • National Institute for Health and Care Excellence (NICE)