Background

In patients with rheumatoid arthritis (RA), problems at work and loss of work may cause socio-economic consequences not only for the patient and their family and the employer but also for the society. For many patients with RA, work is an important part of their life, and being employed increases self-esteem, sense of purpose and financial independence. For employers supporting employees with a chronic disease such as RA, retaining experienced personnel may be more cost-effective than replacing and training new personnel. There has recently been an increase in research of the problems patients with RA may experience with work, including reduced productivity (i.e., presenteeism), sickness absence (absenteeism) and job loss. A better understanding of these problems will enable both patients and employers to intervene early in the disease to reduce possible problems at work, and to prevent sick leave and possible work disability in the long term. Problems at work may be related to the disease itself or to the nature of the working conditions. A particular challenge with RA is that symptoms fluctuate and that it is difficult to be certain about long-term prognosis, which poses difficulties when planning work and commitment to future work tasks. Certain symptoms of RA, including joint damage, pain and fatigue, may not be visible to colleagues and employers, making it sometimes difficult to understand what problems the individual is experiencing and to sympathise with patients when they seek help. On the other hand, many patients will not reveal their disease to colleagues or employers because they are afraid of the consequences this may have on their future employment. This overview gives a brief description of (i) absenteeism and presenteeism in general; (ii) the relationship between the use of biologic therapy and absenteeism, with a focus on sick leave, and on presenteeism; and (iii) indirect costs of absenteeism and presenteeism in those treated with biologic therapies.

Absenteeism and presenteeism

Work productivity can be referred to as the economic productivity of a workplace. It is associated with input, costs, outputs and profits, but it is also dependent upon the quality/quantity and effectiveness of each individual employee (worker productivity) . Workers who have not been lost to the workforce can be less productive as a result of absenteeism and presenteeism. Absenteeism, presenteeism and job loss are interrelated and codependent. For example, rates of presenteeism may be decreased by forcing people to drop completely out of the workforce, but this could not necessarily be described as a ‘good’ thing. Additionally, when workers are pressurised not to take sickness absence, the inevitable consequence is that presenteeism is increased. In studies of the impact of RA on work, different measures of job loss, presenteeism and absenteeism have been employed, making comparison difficult. For example, absenteeism (number of days of sickness absence) can be measured using questionnaires completed by the individual patient, which is subject to both respondent and recall bias. Furthermore, questionnaires sometimes ask people if their days of sick leave were ‘because of their RA’, but this attribution is down to the individual, and respondents may attribute differently. Rates of sickness absence can also be measured by company records or benefit databases, but it is clear that these data are not always collected reliably and coded accurately. Company records are subject to the ‘healthy worker’ bias and benefits databases are prone to misattribution.

To date, where work outcomes have been measured in RA, most of the work has focussed on absenteeism, but there is growing awareness of the socio-economic consequences of presenteeism. Patients with RA in employment may transition between presenteeism, short-term absenteeism and no productivity loss . The pathway between these transitions varies in the same individual over time and between individuals, and it is determined by health status (e.g., flare and remission), personal factors (e.g., age and financial situation) and environmental factors (e.g., job type and company size) ( Fig. 1 ). However, when these transitions have occurred through several cycles, they become less sustainable for the individual and their employer and colleagues. Eventually, this may lead to work disability or to early retirement due to RA.

Transition between presenteeism and absenteeism. Adapted from Tang et al. and Brouwer et al. .

Previous reviews on work disability in RA in Finland, the Netherlands, UK, Canada and USA have shown that 20–70% of patients become work-disabled within 5–10 years after symptom onset, with a 50% probability of becoming work-disabled within 4.5–22 years (median 13 years) after symptom onset . This wide variation in rates between countries may be due to differences in the characteristics of the study populations (e.g., age and gender), social policies, cultural and economic circumstances, physical demands of work and disease severity . In a large study including patients from 32 countries (QUEST-RA), a discrepancy was observed between disease activity in patients from countries with high gross domestic product (GDP) compared with those in low-GDP countries, in that patients from low-GDP countries were more likely to remain working while having higher levels of disability and more severe disease activity . It has also been suggested that rates of work disability have been reducing over the past two decades. This decline might partly be explained by the change in treatment strategies observed during this period, from a wait-and-see approach towards an early and aggressive treatment strategy. However, fluctuations in the general employment market should be taken into account when evaluating changes in employment status over time. In a Swedish study comparing sickness absence between newly diagnosed RA patients 4 years before and 4 years after diagnosis with sickness absence in people from the general population, the mean annual days on sick leave and disability pension increased during the 2 years prior to diagnosis (from 43 to 77 days), and they were highest during the year after diagnosis in the RA population . Thereafter, the mean annual number of days on sickness absence and disability pension remained twice as high as that observed in the general population. This study also compared RA with general population cohorts over three different time periods (year of diagnosis: 1999, 2003 and 2007) to explore period effects. Interestingly, there were similar trends over all three time periods regarding sick-leave days 1 year prior to diagnosis, with the greatest increase in the number of days absent from 1 year prior to 1 year after diagnosis observed in 1999 (89 days, 95% confidence interval (CI): 73–105) and 2003 (75 days, 95% CI: 63–87) compared with 2007 (42 days, 32–51). One year after diagnosis, a steady incline in absent days was observed in both the RA population and general population, whereas the rate remained stable in both the 2007 RA population and general population highlighting the importance of simultaneous measuring temporal trends in absenteeism in the general population when presenting changes in absenteeism over time in RA populations .

Instruments to measure absenteeism and presenteeism

A number of questionnaires have been developed to assess absenteeism with a focus on the number of days absent due to ill health, number of days on sick leave and job loss (i.e., work disability or early retirement) . Not all instruments are validated, and many lack information on reliability. Measuring presenteeism is even more complicated as many factors have an impact on the output, including job type, control over work, replacement by colleagues, size of company, profession and financial situation . Moreover, the output of a job is not always based on quantity, but it could be based on quality or both . In a job producing goods, the number of goods produced may be reduced due to RA as a patient may need more time to handle goods or has to take more breaks during the day. For a teacher with RA, the quality of work may be reduced due to fatigue. To truly measure the value of presenteeism in these situations thus remains difficult.

To overcome some of these issues, a number of presenteeism instruments have been developed, which can be used in different settings including employees of the general population and/or patients with RA in paid employment . Although all measures were developed to value ‘presenteeism’, differences exist in the construct, references used, recall period and whether the measurement is generic or disease specific . The content of a measure includes perceived impairment, productivity, performance, efficiency and quantity and quality. Generic measures are useful when comparing the impact of RA with other (chronic) diseases, whilst disease-specific measures may be more desirable in clinical studies including patients with RA.

Absenteeism and presenteeism

Work productivity can be referred to as the economic productivity of a workplace. It is associated with input, costs, outputs and profits, but it is also dependent upon the quality/quantity and effectiveness of each individual employee (worker productivity) . Workers who have not been lost to the workforce can be less productive as a result of absenteeism and presenteeism. Absenteeism, presenteeism and job loss are interrelated and codependent. For example, rates of presenteeism may be decreased by forcing people to drop completely out of the workforce, but this could not necessarily be described as a ‘good’ thing. Additionally, when workers are pressurised not to take sickness absence, the inevitable consequence is that presenteeism is increased. In studies of the impact of RA on work, different measures of job loss, presenteeism and absenteeism have been employed, making comparison difficult. For example, absenteeism (number of days of sickness absence) can be measured using questionnaires completed by the individual patient, which is subject to both respondent and recall bias. Furthermore, questionnaires sometimes ask people if their days of sick leave were ‘because of their RA’, but this attribution is down to the individual, and respondents may attribute differently. Rates of sickness absence can also be measured by company records or benefit databases, but it is clear that these data are not always collected reliably and coded accurately. Company records are subject to the ‘healthy worker’ bias and benefits databases are prone to misattribution.

To date, where work outcomes have been measured in RA, most of the work has focussed on absenteeism, but there is growing awareness of the socio-economic consequences of presenteeism. Patients with RA in employment may transition between presenteeism, short-term absenteeism and no productivity loss . The pathway between these transitions varies in the same individual over time and between individuals, and it is determined by health status (e.g., flare and remission), personal factors (e.g., age and financial situation) and environmental factors (e.g., job type and company size) ( Fig. 1 ). However, when these transitions have occurred through several cycles, they become less sustainable for the individual and their employer and colleagues. Eventually, this may lead to work disability or to early retirement due to RA.

Transition between presenteeism and absenteeism. Adapted from Tang et al. and Brouwer et al. .

Previous reviews on work disability in RA in Finland, the Netherlands, UK, Canada and USA have shown that 20–70% of patients become work-disabled within 5–10 years after symptom onset, with a 50% probability of becoming work-disabled within 4.5–22 years (median 13 years) after symptom onset . This wide variation in rates between countries may be due to differences in the characteristics of the study populations (e.g., age and gender), social policies, cultural and economic circumstances, physical demands of work and disease severity . In a large study including patients from 32 countries (QUEST-RA), a discrepancy was observed between disease activity in patients from countries with high gross domestic product (GDP) compared with those in low-GDP countries, in that patients from low-GDP countries were more likely to remain working while having higher levels of disability and more severe disease activity . It has also been suggested that rates of work disability have been reducing over the past two decades. This decline might partly be explained by the change in treatment strategies observed during this period, from a wait-and-see approach towards an early and aggressive treatment strategy. However, fluctuations in the general employment market should be taken into account when evaluating changes in employment status over time. In a Swedish study comparing sickness absence between newly diagnosed RA patients 4 years before and 4 years after diagnosis with sickness absence in people from the general population, the mean annual days on sick leave and disability pension increased during the 2 years prior to diagnosis (from 43 to 77 days), and they were highest during the year after diagnosis in the RA population . Thereafter, the mean annual number of days on sickness absence and disability pension remained twice as high as that observed in the general population. This study also compared RA with general population cohorts over three different time periods (year of diagnosis: 1999, 2003 and 2007) to explore period effects. Interestingly, there were similar trends over all three time periods regarding sick-leave days 1 year prior to diagnosis, with the greatest increase in the number of days absent from 1 year prior to 1 year after diagnosis observed in 1999 (89 days, 95% confidence interval (CI): 73–105) and 2003 (75 days, 95% CI: 63–87) compared with 2007 (42 days, 32–51). One year after diagnosis, a steady incline in absent days was observed in both the RA population and general population, whereas the rate remained stable in both the 2007 RA population and general population highlighting the importance of simultaneous measuring temporal trends in absenteeism in the general population when presenting changes in absenteeism over time in RA populations .

Instruments to measure absenteeism and presenteeism

A number of questionnaires have been developed to assess absenteeism with a focus on the number of days absent due to ill health, number of days on sick leave and job loss (i.e., work disability or early retirement) . Not all instruments are validated, and many lack information on reliability. Measuring presenteeism is even more complicated as many factors have an impact on the output, including job type, control over work, replacement by colleagues, size of company, profession and financial situation . Moreover, the output of a job is not always based on quantity, but it could be based on quality or both . In a job producing goods, the number of goods produced may be reduced due to RA as a patient may need more time to handle goods or has to take more breaks during the day. For a teacher with RA, the quality of work may be reduced due to fatigue. To truly measure the value of presenteeism in these situations thus remains difficult.

To overcome some of these issues, a number of presenteeism instruments have been developed, which can be used in different settings including employees of the general population and/or patients with RA in paid employment . Although all measures were developed to value ‘presenteeism’, differences exist in the construct, references used, recall period and whether the measurement is generic or disease specific . The content of a measure includes perceived impairment, productivity, performance, efficiency and quantity and quality. Generic measures are useful when comparing the impact of RA with other (chronic) diseases, whilst disease-specific measures may be more desirable in clinical studies including patients with RA.

Absenteeism in patients treated with biologics

Since the introduction of etanercept, many new biologic therapies have been approved to treat RA showing good results in controlling the disease and preventing radiographic progression in most patients. However, in most countries, biologic therapies can only be prescribed after the failure of two synthetic disease-modifying anti-rheumatic dugs (sDMARDs) by which time many patients may have already lost their job . Given the change in treatment strategies to early aggressive treatment and the increased availability of biologic therapies in the past decade, and the accessibility of biosimilars in the near future, delays between diagnosis and commencement of a biologic (if one is required) have decreased, and they will probably further decrease during the next few years, pending health-care budget limitations. Early intervention, together with better job management, is important for retaining paid work as return to work is difficult for those who are work-disabled . To date, however, the influence of biologics on absenteeism has mainly been investigated in patient populations with long-standing RA.

The studies exploring the effect of biologic therapy on employment status have produced conflicting results , which can partly be explained by differences in study design (i.e., randomised clinical trial (RCT), open-label extension RCT or observational cohort study), inclusion of patients with early or established disease, instrument used to measure employment status, when the assessment of employment status was measured (i.e., prior and after start biologic therapy, at biologic therapy commencement and during follow-up after biologic therapy), follow-up duration, recall period and selection of patient population. For example, in the British Society for Rheumatology (BSR) Biologics Register for RA in the UK (BSRBR-RA), an observational study including patients starting biologic therapy and a control group of patients on sDMARDs, employment status did not differ between patients with RA receiving biologic therapies compared with patients using sDMARDs . Those patients who responded to anti-tumour necrosis factor (TNF) therapy were, however, less likely to become work-disabled. In the PREMIER study, an RCT comparing adalimumab plus methotrexate (MTX) with MTX alone, patients receiving the combination therapy were 1.53 (95% CI: 1.04–2.56) times more likely to retain/regain employment .

Absenteeism includes, in addition to work loss, the number of days on sick leave due to RA. Table 1 gives an overview of studies evaluating days of sick leave in patients with RA using biologic therapies including 10 RCTs , three open-label RCTs and nine observational studies . Even though the description of absenteeism seems relatively straightforward, a number of different definitions were used to define the number of days absent from work in these studies. A few studies used validated questionnaires to estimate the number of days absent including the Work Productivity Survey for Rheumatoid Arthritis (WPS-RA) , Health Labour Questionnaire (HLQ) and Work Productivity and Activity Impairment questionnaire (WPAI-RA) . In two studies, Short Form (SF)36 or Health Assessment Questionnaire (HAQ) scores were mapped onto existing measures of presenteeism (i.e., WPAI and Work Limitation Questionnaire (WLQ)) using regression techniques to estimate the inability to work and the number of days missed due to absenteeism/presenteeism , whereas a zero-inflated Poisson (ZIP) model, which accounts for excessive zero values in absenteeism, was used in another study . In patients with early RA (a disease duration of <2 years), a significant decrease in absent workdays was observed after commencement of biologic therapy , except in the study by Eriksson et al. .

Table 1

Effect of biologic treatment on absenteeism.

Author	Study design	Treatment	N	Disease duration	Absenteeism measure	Time points	Result	Sign.
RCTs
Smolen, 2006	RCT (ASPIRE)	MTX	282	0.9 (0.7) ∗ years	Number of days missed from work (0 days, 1–10 days or >10 days)	Baseline – 54 weeks	% patients MTX vs. INF + MTX:	[↑]
		INF + MTX	722				0 days: 66.6% vs. 78.9%
							1–10 days: 16.7% vs. 11.5%
							>10 days: 16.7% vs. 9.6%
Cole, 2008	Phase III RCT (AIM and ATTAIN)	ABA (MTX failure sample)	652	NA	PCS-based estimates inability to work	Baseline, 6 months, 1 and 2 years	PCS-based: AIM, mean (SD): 15.5 (13.0) vs. 15.5 (13.5)	[ = ]
		ABA (anti-TNF failure sample)	391		MCS-based estimates inability to work		ATTAIN, mean (SD): 21.9 (17.4) vs. 20.2 (13.6)	[ = ]
							MCS-based:AIM, mean (SD): 1.9 (1.2) vs. 2.0 (1.2)	[ = ]
							ATTAIN, mean (SD): 2.1 (1.4) vs. 1.8 (1.2)	[ = ]
Bejarano, 2008	RCT	ADA + MTX	75	9.5 (6.0) ∗ months	WIS RA score	Baseline – 56 weeks	ΔWIS RA MTX vs. ADA + MTX: −5.4 (−7.7) vs. −8.1 (6.9) Δ −2.7 (95% CI: −5.1 to −0.4)	[↑]
		MTX	73	7.9 (5.4) ∗ months	Number of working days lost		% work time lost: 18.4 (34.1) vs. 8.6 (23.3), Δ −9.9 (95% CI: −19.2 to −0.5)	[↑]
Kavanaugh, 2009	Phase II RCT (RAPID I and RAPID II)	CZP200 + MTX, CZP400 + MTX, placebo	162	6.1 (4.2) ∗ years	WPS-RA absenteeism: full workdays missed due to RA	0, 4, 24 and 52 weeks (RAPID I only)	RAPID I:
			139	6.2 (4.4) ∗ years			MTX, mean: 4.6, 2.5, 4.4, 4.5	[ = ]
			69	6.2 (4.4) ∗ years			CZP200 + MTX, mean: 3.1, 1.5, 1.0, 1.0	[↑] 24 and 52 weeks
			101	6.1 (4.1) ∗ years			CZP400 + MTX, mean: 4.5, 1.2, 1.5, 1.4
		CZP200 + MTX, CZP400 + MTX, placebo	95	6.5 (4.3) ∗ years			RAPID II:
			49	5.6 (3.9) ∗ years			MTX, mean: 3.01, 2.4, 2.5	[ = ]
							CZP200 + MTX, mean: 3.8, 2.1, 2.3	[ = ]
							CZP400 + MTX, mean: 2.8, 1.9, 1.0	[ = ]
Anis, 2009	RCT (COMET)	MTX	100	8.9 (5.6) months	Number of missed workdays	0 and 12 months	Missed workdays MTX vs. ETN + MTX, bootstrapped mean (95% CI): 31.9 vs. 14.2, Δ −17.6 (95% CI: −34.4, −2.2)	[↑]
		ETN + MTX	105	8.6 (5.3) months	Reduced working time (in days)		Reduced working time MTX vs. ETN + MTX, bootstrapped mean (95% CI): 19.8 vs. 10.5, Δ −9.3 (95% CI: −21.9, 3.9)	[ = ]
					Number of stopped workdays ∗∗∗		Stopped worked days MTX vs. ETN + MTX, bootstrapped mean (95% CI): 32.9 vs. 10.9, Δ −22.1 (95% CI: −45.2, −0.3)	[↑]
van Vollenhoven, 2010	RCT (PREMIER)	MTX	214	0.8 (0.9) ∗ years	No. days missed (estimated using ZIP model):	0 and 24 months	Comb. vs. MTX, total: 17.4 vs. 36.9	[↑]
		ADA + MTX	219	0.8 (0.8) ∗ years			ADA vs. MTX, total: 18.7 vs. 36.9	[↑]
		ADA + placebo	231	0.7 (0.8) ∗ years
Hazes, 2010	Phase III RCT (RAPID I and RAPID II)	Pooled data: CZP200/400 + MTX	982	6.1 (4.3) ∗ years	WPS-RA, missed days in last month	Baseline, 12 weeks: HAQ-DI resp. vs. non-resp.	Resp. vs. non-resp., mean change: −2.1 vs. −1.8 days	[ = ]
			19	6.2 (4.2) ∗ years
Zhang, 2012	RCT (COMET)	ETN + MTX or MTX (total group)	214	8.7 months	Missed days from work	0–12, 12–24, 24–36 and 36–52 weeks	Mean (SD): 0.61 (1.51), 0.41 (1.15), 0.29 (0.91), 0.20 (0.72)
							18.5%, 18.8%, 16.5%, 14.8%
					Work absence		ACR20–50 resp. vs. no resp.: adj OR (95% CI): 1.08 (0.52–2.27)	[ = ]
							ACR50–70 resp. vs. no resp.: 0.87 (0.43–1.78)	[ = ]
							ACR70 vs. no resp.: 0.45 (0.21–0.98)	[↑]
							DAS28 2.6–3.2 vs. >3.2: 0.98 (0.46–2.12)	[ = ]
							DAS28 < 2.6 vs. >3.2: 0.79 (0.43–1.43)	[ = ]
Eriksson, 2014	RCT (Swefot)	INF	128	10.5 (3.5) months	Sick-leave days	Baseline, 12 and 21 months	Mean change over 12 months INF vs. conv.: −4.1 vs. −4.0 (adj Δ 0.1, 95% CI: −2.7–3.3)	[ = ]
		Conventional treatment	130	10.0 (3.1) months			Mean change over 12 months INF vs. conv.: −4.9 vs. −6.2 (adj Δ 0.1, 95% CI: −1.2–4.4)	[ = ]
Allaart, 2007	RCT (BeSt)	I = Seq. monotherapy	NA	NA	Three-monthly diary on work absenteeism	Baseline until 2 years	Overall: decrease of 0.1 on utility associated with decrease of 2 working h/week	[?]
		II = step-up comb. therapy
		III = initial comb. therapy					Productivity was highest in group IV
		Iv = MTX + INF
Open-label extension RCT
Zhang, 2008	open-label Phase IIIb RCT (CanAct trial)	ADA	140	10.1 (8.1) years	HLQ: no. of absent workdays	0 and 12 weeks	No. absent days, mean (SD): 1.3 (3.1) to 0.8 (2.6), Δ −0.5 (2.9), (95% CI: −1.0, −0.0) ∗∗	[↑]
						ACR20 resp. vs. non-resp.	−0.3 (2.4) vs. −0.5 (3.1), (95% CI: −0.7, 1.1)	[ = ]
Mittendorf, 2008	Open-label extension study	ADA	144	12.4 (7.7) ∗ years	Sick leave over 6 months	−6 months prior treatment vs. +6 months during	Days on sick leave, mean (SD): 12.6 (SD 33.9) vs. 8.35 (SD; 16.8)	[?]
Pavelka, 2013	Open label (PRESERVE) (CEE)	ETN + MTX	150	6.0 (6.1) ∗ years	WPAI-RA: per cent of work time missed due to RA	0 and 36 weeks	Work time missed, mean (SD): 11.82 (25.78), 2.43 (10.64), Δ median per cent −100.0 (−100, −94)
Observational studies
Yelin, 2003	Observational study	ETN (RAPANEL)	259	17.2 (8.8) ∗ years	Adjusted mean number of work hours per week.	Over 12-month period prior to last interview	h/week RAPANEL vs. RAPOLO, adj mean: 38.0 vs. 38.5, Δ −0.5 (95% CI −4.0–5.0)	[ = ]
		Non-ETN (RAPOLO)	238	14.0 (8.3) ∗ years	Adjusted mean number of weeks per year.		weeks/year RAPANEL vs. RAPOLO, adj mean: 46.8 vs. 48.6, Δ −1.8 (95% CI −4.7–1.2)	[ = ]
Kobelt, 2004	SSATG register	INF, ETN	116	14.2 (9.1) years	Number of days on sick leave	−12 months and +12 months	Mean (SD) −12 months vs. +12 months: 1.6 (5.0) vs. 1.1 (2.6)	[↑]
Laas, 2006	Observational	INF	51	16 (3–43) ∗ years	Days off work	−12 months and +12 months	Period I vs. period II, mean: 121 vs. 141	[?]
Farahani, 2006	Cohort study	ETN	223	12.5 (SD 9.2) ∗ years	Missed days of work	Baseline, 6 and 12 months	Baseline –6 months ETN vs. no ETN, mean (SD): 2.5 (7) vs. 7.8 (19)	[↑]
		non-ETN, eligible	208	12.3 (SD 9.7) ∗ years			Baseline –12 months ETN vs. no ETN, mean (SD): 6.0 (14) vs. 7.6 (18)	[ = ]
Olofsson, 2011	SSATG Register	anti-TNF	365	55.4 (23.4–124.6) months	Days of sick leave 12 months prior and after start anti-TNF	−12 and + 12 months	Mean (95% CI): 9.3 (95%CI 8.0–10.6) vs. 6.6 (95%CI 5.5–7.8)	[↑]
Augustsson, 2010	Observational (STURE register)	anti-TNF	594	9.4 (8.5) years	Hours worked/week	Baseline, 6 months, 1, 2, 3, 4 and 5 years	Unadjusted model: improvement 4.2 h/week in first year, 0.5/h thereafter	[↑]
							Adjusted model: improvement 4.1 h/week first week, no change thereafter	[↑]
Krüger, 2012	Non-interventional study	ADA	1157	9.3 (7.5) years	Weeks of sickness absence	−12 and + 12 months	Mean (SD) days: 4.65 (SD 9.1) vs. 4.52 (SD ?)	[?]
Nair, 2013	Panel survey database	non-users:		NA	Half day or more days of work missed	Cross-sectional	Adj IRR (95% CI) (non-users with mild RA = reference):
		mild RA	720				1.92 (95% CI: 0.97–3.79	[ = ]
		moderate RA	159				1.99 (95% CI: 0.35–11.37	[ = ]
		severe RA	208				0.61 (95% CI: 0.23–1.58)	[ = ]
		anti-TNF alpha	65
Hone, 2013	Prospective observational study	ETN	N = 196 (All)	5.4 (8.1) years	WPAI-RA: absenteeism score work time missed	0 and 6 months	All (%): 9.9 (20.1) to 4.4 (16.1), Δ mean change −3.5 (17.0)	[↑]
			N = 147 (Continuers on ETN)				Cont.: 8.4 (18.8) to 2.3 (9.9), Δ mean change −4.1 (14.2)	[↑]

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:

Psychological and psychosocial determinants of musculoskeletal pain and associated disability Ageing, musculoskeletal health and work Shoulder disorders and occupation The impact of ankylosing spondylitis/axial spondyloarthritis on work productivity Rheumatic effects of vibration at work Economic impact of musculoskeletal disorders (MSDs) on work in Europe

Stay updated, free articles. Join our Telegram channel

Join