Discussion

Overall, the rate of sickness certification for back pain has declined significantly between 2000 and 2010. This decline in rates has principally occurred from 2003 onwards. This study demonstrated that approximately one-third of consultations for back pain will lead to a sickness certificate; this finding is in line with many studies examining back pain which also find that approximately one-third of participants are absent from work.14 ,15

There was a trend in certification by gender, with men reporting consistently higher rates of certification than women. This is to be expected and has been reported in previous studies of certification using this data set.12 ,16 The most likely explanation is the increased proportions of males working in the manual sector at 34.6% when compared to females at 10.9%.17 On average, manual workers report more absence than non-manual employees, and also report more absence as a result of musculoskeletal conditions than non-manual employees.5 The differences in certification for gender may also be associated with differing consultation behaviours, and it has been reported that men are more likely to consult with an episode of back pain than women at a female:male rate ratio of 1:20 (95% CI 1.13 to 1.28).13

There was no clear trend in certification by age with the exception of the 60+ years age group, which consistently had a lower rate of certification compared to younger age groups. Again, the rate of certification in this group showed a downward trend, but it was not as pronounced as in other groups. There are a number of reasons why this difference may be seen in the 60+ age group. First, there are fewer people employed in this age group; in 2011, the Organisation for Economic Co-operation and Development reported that 56% of the population in the 55–64 years age group were employed compared to 80% of the population in the 25–54 years age group.18 Second, data from 2013 demonstrated that while individuals in the 50–64 years age bracket are less likely to work on a part-time basis (just 28% of all workers in this age group), those in the over 64 years age group are more likely to report that they work part time, a total of 66% of all those in this age group.19 Third, there could be a healthy worker effect whereby individuals suffering with back pain are removing themselves from the workplace. Findings are not likely to be related to differential reporting, as reporting of back pain over time is likely to be consistent.20

It is difficult to identify the reason for the decreasing rate of sickness certification in this study. It could be argued that the prevalence of back pain has decreased over time and so there is less need for certification. However, the literature suggests that the prevalence of back pain has remained largely unchanged over the period 1990–201021 and may even have increased.22 ,23 It could also be that the number of consultations for back pain has increased; however, the literature again suggests that this is not the case.13 ,24 ,25 During the period of analysis that this paper spans, there have been a number of initiatives, in the UK and worldwide, surrounding the management of back pain that may have contributed to the decline in rates of certification. A review of clinical guidelines for the management of non-specific back pain assessed guidelines published between 2000 and 2008;26 this review is an update of a previous review.27 The authors report that the most common advice is to reassure patients and encourage them to remain active; however, in contrast to the earlier review, the current guidelines increasingly mention return to work, despite back pain, in their recommendations. In the UK, 2009 saw the publication of two National Institute for Clinical Excellence (NICE) reports, the first considering early management of non-specific back pain28 and the second reporting on primary care management of long-term sickness absence.29 The non-specific back pain guidelines28 concur with the review by Koes et al27 that individuals should be encouraged to remain active and continue normal activities as far as possible. The second report is focused on promoting the benefits of working with health conditions,29 specifically back pain, and again encourages individuals to maintain work despite pain. However, the impact of these recent reports is unlikely to be seen in the current analysis.

Recent research from the Global Burden of Disease study has demonstrated that back pain leads to more years lived with disability than any other condition and that this burden of back pain is increasing with an ageing population.21 There is also evidence that the number of consultations for back pain has remained static, at least in the population of the USA, for a period of over 10 years from the 1990s to the mid-2000s.30 It is interesting then to note that, despite the relative stability of consultations for back pain and the relatively adverse outcomes in terms of the burden of back pain, the number of sickness certificates for back pain has decreased in this study. Within the UK, there have not been any recent public health initiatives which may account for this decrease in certification; there has, however, been a sea change in the information that GPs are advised to give their patients in regard to working with pain. There is consistent information provided by the Royal College of General Practitioners,31 Department for Work and Pensions32 and NICE,28 ,29 coupled with the availability of information booklets detailing the management of health and work for the GP, patient and employer.33–36 It may be that media campaigns to promote working with musculoskeletal pain could see the rate of sickness certification reduce further. There has been some success with media campaigns in Scotland and Australia demonstrating that positive messages around working with musculoskeletal pain led to improvements in knowledge, attitudes and beliefs at the population level which were maintained, albeit at a reduced level, 3 years later.37–39 A Canadian study also utilising a media campaign had a more limited effect on behaviours related to back pain, for example, healthcare use, than those run in Scotland and Australia.40 However, the Canadian study did find that participants agreement with the statement ‘if you have back pain you should try to stay active’ significantly increased.40 It seems then that knowledge does not necessarily translate into behaviour and it has also been reported that provision of information alone is not sufficient to prevent work absence,41 but it seems that improving the baseline understanding of the general population may enhance any further information provided by healthcare professionals.42

There are a number of limitations to the current study. First, capturing the duration of sickness absence is not possible in the current data set, meaning it is unclear whether the data are unduly influenced by a large number of short-term or long-term certificates. The data presented are a ‘pure’ rate of certification, that is, a rate of certification per consultation. The data do not take into account the potential for a single patient to be issued multiple certificates for one episode of back pain. The possibility that the rate is artificially inflated as a result of a few patients receiving multiple certificates or deflated as a result of multiple patients receiving one certificate of a long duration cannot be ruled out. However, the most frequently recorded duration for absence is 2 weeks,43–46 and there is no reason to believe that GPs in the current study would differ to any great degree. It is also important to acknowledge that for approximately two-thirds of people with an episode of back pain, a return to work within 1 month is expected,15 indicating that the majority of consultations will be for individual episodes of back pain and therefore be individual certificates.

Second this data set is based in one area of the UK, North Staffordshire, and it could be argued that it is not generalisable to the rest of the population. Previous work with this data set has demonstrated that crude rates of certification change very little when the data are standardised to the age and gender of the population as a whole, and there is no indication that this should be any different for this study.12 Lastly, the data included in the manuscript only goes as far as 2010 as from this date onwards the Read coding for sickness certificates changed as a result of the introduction of the Fit Note. This change in Read coding means that Fit Notes are now coded as ‘not fit for work’ and ‘may be fit for some work’. Since this classification is different from the pre-2010 sickness certificates, it would make comparison of the data between these two periods difficult in this manuscript because we would not be comparing like with like. As a result, we could not be confident that any changes in the rate of certification would be as a result of the change to the Fit Note system or as a result of the change in Read coding, that is, more read codes available to classify not fit for work and may be fit for some work.

The main strength of this study is the large database in which this work was carried out. This data set allowed the linking of consultation data to certification data to enable trends by age and gender to be seen. This is the first study to map trends in certification for back pain in the UK using an established data set which has been validated for use in examining both consultation and sickness certification data.11 ,13 ,47 There is a need for establishing baseline rates of certification against which any change in policy or strategy at either the local or national level can be compared, and this data set goes some way to establishing these baseline figures. Linaker et al48 state that improvement of existing data sets would be more attractive than the development of a new national system to record sickness absence; with the introduction of the e-Fit Note, the current data set is being updated to include information on duration of absence and whether or not a patient may be fit for work, further strengthening this data set for future research. Another strong point of this study is that is appears to be supported by other literature; this lends credibility to our findings. For example, Ruseckaite et al49 reported a significant decrease in certification for musculoskeletal disorders between 2003 and 2010 in Australia. They also report a decrease in the rate of certification for back pain and strains, although there is no indication of whether this trend is significant or not. Gabbay et al50 also reported a decrease in the number of Fit Notes, lasting over 12 weeks, issued over a 12-month period. Lambeek et al51 reported that the costs of back pain in the Netherlands decreased between 2002 and 2007, acknowledging that a large proportion of this cost is made up of absence. In the UK, the Office for National Statistics has reported a decrease in reported absence across all health conditions between 1993 and 2013.52 However, work still needs to be carried out to investigate why there has been this change, and whether it is truly related to the available evidence about working with pain.