Longitudinal follow-up of employment status in patients with CFS after mononucleosis (Nyland et al.)

Dolphin

Senior Member
Messages
17,557
Likes
28,249
Free full text at: http://bmjopen.bmj.com/content/4/11/e005798.long

For anyone interested, one can also read the review history here: http://bmjopen.bmj.com/content/4/11/e005798.reviewer-comments.pdf and the earlier drafts here: http://bmjopen.bmj.com/content/4/11/e005798.draft-revisions.pdf

BMJ Open 2014;4:e005798 doi:10.1136/bmjopen-2014-005798
  • Health economics
Longitudinal follow-up of employment status in patients with chronic fatigue syndrome after mononucleosis
  1. Morten Nyland1,
  2. Halvor Naess1,2,
  3. Jon Steinar Birkeland2,
  4. Harald Nyland1
+Author Affiliations

  1. haln@haukeland.no
  • Received 28 May 2014
  • Revised 31 October 2014
  • Accepted 6 November 2014
  • Published 26 November 2014
Abstract
Objective To examine the effect of early clinical and demographic factors on occupational outcome, return to work or awarded permanent disability pension in young patients with chronic fatigue syndrome (CFS).

Design Longitudinal cohort study.

Intervention A written self-management programme including a description of active coping strategies for daily life was provided.

Setting, participants Patients with CFS after mononucleosis were evaluated at Department of Neurology, Haukeland University Hospital during 1996–2006 (contact 1). In 2009 self-report questionnaires were sent to all patients (contact 2).

Primary and secondary outcome measures Primary measure was employment status at contact 2. Secondary measures included clinical symptoms, and Fatigue Severity Scale (FSS) scores on both contacts, and Work and Social Adjustment Scale (WSAS) at contact 2.

Results Of 111 patients at contact 1, 92 (83%) patients returned the questionnaire at contact 2. Mean disease duration at contact 1 was 4.7 years and at contact 2 11.4 years. At contact 1, 9 (10%) were part-time or full-time employed. At contact 2, 49 (55%) were part-time or full-time employed. Logical regression analysis showed that FSS≥5 at contact 2 was associated with depression, arthralgia and long disease duration (all at contact 1).

Conclusions About half of younger patients with CFS with long-term incapacity for work experienced marked improvement including full-time or part-time employment showing better outcomes than expected. Risk factors for transition to permanent disability were depression, arthralgia and disease duration.
Strengths and limitations of this study
  • Two strengths of the study are very long prospective follow-up period and focus on employment.

  • A limitation is that patients were recruited from a tertiary centre.

  • Long-term prognosis for young patients with chronic fatigue syndrome after mononucleosis is favourable for a large subgroup.

  • More than half of the patients with long-term incapacity for work are re-employed after mean disease duration of 11.4 years.

  • Factors associated with poor long-term prognosis include depression, arthralgia and disease duration.
 

Dolphin

Senior Member
Messages
17,557
Likes
28,249
An important thing to note is this was a young cohort
[text says mean age of 23 but the text also says "the mean age of the patients at the onset of CFS was 23.7 years (SD=7.3)" and Table 2 (which is the group that they actually have data for) suggests 24].
[To be exact: one group's mean (SD) age was 23.8 (7.9) (n=36); for the other group it was 24.1 (7.0)]
ETA: I misread this slightly. This is the age at onset. The age at the first assessment was 26.8 (7.5) for one group (n=36) and 29.3 (7.0) for another group (n=54).

Studies have suggested younger people have a better prognosis.

---
The abstract says 10% were in part-time or full-time employment at baseline.
It doesn't mention: "12 patients (13.5%) were students".
At time point two, "one patient (1%) was a student."

Related to this:
At the time of mononucleosis 43 (47%) were employed at work and 48 (52%) were students (missing data in one patient).
So one might say that part of the reported increase in employment is somewhat artificial as being a student can be as demanding as working.

----

Also, patients with mono may have a better prognosis than other forms of CFS. For example, in Katz et al. 13% of adolescents had CFS at 6 months but only 4% had CFS at 24 months. That means that 9/13 (69%) of those who had CFS at 6 months did not have CFS at 24 months.

The total cohort the authors could have written to was 873 so this is only 12.7% of them. They don't say if they wrote to the others or not. This group may not be representative.
 
Last edited:

A.B.

Senior Member
Messages
3,780
Likes
23,077
CFS following mononucleosis probably has a better prognosis than other forms. Unfortunately that's a minority:
The 111 patients constitute all patients diagnosed with CSF triggered by mononucleosis in the total cohort of 873 patients.
What is happening with the other 762 patients?

I also like this part (and I'm sure Mr Wessely does not)

We found that all patients who were unemployed at the initial examination received sickness or disability benefits. Norway has been criticised for high-disability payments which may undermine motivation for individuals to stay in work.31 A poor response to treatment for CFS was predicted by being in receipt of sickness benefits in a patient education study.28 In contrast, this study shows that long-term compensations to secure the socioeconomic position does not inhibit return to work, but may be essential contributors to the high proportion becoming employed at final follow-up. In addition to the financial support the contact with the social security system initiates rehabilitation activities directed towards obtaining new work when unemployed.18
 
Last edited:

Dolphin

Senior Member
Messages
17,557
Likes
28,249
At contact 2 24 (27%) were fully employed, 25 (28%) were employed part-time and 40 (45%) were unemployed (missing data in three patients). One patient (1%) was a student. In total, 63 of 92 patients received DP or sickness absence benefits: 15 patients (17%) were awarded partial DP and 39 (44%) received full DP for the reduced working capacity, 6 patients (7%) got partial SA benefits and 3 patients (3%) full SA benefits. One (1%) unemployed patient was part-time student. Five (5%) patients were employed at both contact 1 and contact 2.Figure 1shows employment status at contact 1 and contact 2.
 

Esther12

Senior Member
Messages
13,774
Likes
28,358
Thanks.

I only had a very brief look, but this seems a bit too positive:

Our main finding was that about half of the patients improved during the study period and were fully or partly employed at the final follow-up.
The change in the number unemployed was from 70% to 40%:

http://bmjopen.bmj.com/content/4/11/e005798/F1.large.jpg

(While I was typing Dolphin has been pointing out similar things).
 

Esther12

Senior Member
Messages
13,774
Likes
28,358
The don't cite the PACE employment data... I wonder if it being presented in the cost-effectiveness paper means lots of researchers just aren't aware of it?

A few longitudinal studies have reported employment at baseline and follow-up after intervention. A long-term study of cognitive behaviour therapy versus relaxation therapy evaluated outcome at 5-year follow-up. A total of 68% of the 25 patients who received cognitive therapy rated themselves as improved compared to 36% of the 28 patients who received relaxation therapy. Similar proportions of patients were employed (56% vs 39%) but the patients in the cognitive behaviour group worked more hours per week (36 vs 24).26 In another study no treatment effect of cognitive behaviour therapy as compared with natural course was found on work rehabilitation although self-rated improvement was associated with cognitive behaviour treatment.27

A randomised controlled trial of patient education to encourage graded exercise resulted in substantial self-reported improvement in physical and occupational functioning compared with standard medical care. The receipt of sickness benefit at the start of treatment was associated with poor outcome.28 Occupational therapy with a lifestyle management programme was offered to 74 patients after median illness duration of 5 years. At follow-up 18 months later 31 (42%) of the patients had returned to new employment, voluntary work or training.29
 

Dolphin

Senior Member
Messages
17,557
Likes
28,249
The abstract and summary points both give the impression that depression was associated with employment outcomes.

However, the text makes this clear this is not the case:

We found that depression at the first contact tended to predict poor prognosis both as to FSS and WSAS scores, but not employment.
This is true both for depression measured at the first timepoint and also at the second timepoint.

Here's the WSAS questionnaire:
Please read each of the following questions and tick the appropriate box to indicate

No impairment Severe impairment
0 1 2 3 4 5 6 7 8

Because of my CFS/ME, my ability to work is Impaired.

Because of my CFS/ME, my social leisure activities (with other people, such as parties, bars, clubs, outings. visits, dating, home entertainment) are impaired.

Because of my CFS/ME, my home management (cleaning, tidying, shopping, cooking, looking after home or children, paying bills) is impaired.

Because of my CFS/ME, my private leisure activities (done alone, such as reading, gardening. collecting, sewing, walking alone) are impaired.

Because of my CFS/ME, my ability to form and maintain close relationships with others, including those I live with, is impaired.
Here's the FSS:
Table 2—Fatigue Severity Scale (FSS)'

Statement
1. My motivation is lower when I am fatigued.
2. Exercise brings on my fatigue.
3. I am easily fatigued.
4. Fatigue interferes with my physical functioning.
5. Fatigue causes frequent problems for me.
6. My fatigue prevents sustained physical functioning.
7. Fatigue interferes with carrying out certain duties and responsibilities.
8. Fatigue is among my three most disabling symptoms.
9. Fatigue interferes with my work, family, or social life.

'Patients are instructed to choose a number from 1 to 7 that indicates their degree of agreement with each statement where 1 indicates strongly disagree and 7, strongly agree.
 
Last edited:

Simon

Senior Member
Messages
3,788
Likes
14,822
Location
Monmouth, UK
Just to clarify, for my own benefit as much as anything.
  • This is a study of CFS patients whose illness began with mono/glandular fever and who had been ill for an average of 4.7 years at baseline and at that point had an average age of 28.
  • Follow-up was at varying times and patients were ill for an average of 11.4 years at follow-up (does that include those who rated themselves as recovered?)
  • 70% were unemployed at baseline, 40% at follow up, a 43% fall, (but nb for a healthy population the unemployment rate might not be zero eg could be 10%, esp if it includes people not working through choice eg raising families)
  • Based on self-assessment, 70% had improved, including 13% recovered (though does that include minor improvement?)
I'm not sure what to make of these results. The improvement rates look much better than I seem to recall from elsewhere given they had already been ill for an average of five years, but they were quite a young group. It's also possible that most of the improvement came from those who had been ill for a shorter time at baseline, since illness duration at baseline correlated with fatigue at follow-up.
 

alex3619

Senior Member
Messages
13,810
Likes
37,668
Location
Logan, Queensland, Australia
We know from the Dubbo studies, and Lloyd's recent comment reported on the Simmaron blog, that only about 2% of EBV induced CFS go on to near permanent illness. Most recover, though slowly. How do we know this does not reflect the natural time course? I actually consider it a dismal result.

Edit: correction, that's 2% of EBV patients, or about 20% of CFS-like post mono patients.
 
Last edited:
Messages
5,902
Likes
12,696
Location
South Australia
The 'recovery' rates are still quite low, what we have here is ill people who are still managing to work.

One alternative take home message, is that all of the various laws towards non-discrimination towards those with disabilities, providing greater accessibility, supportive working environments etc may actually be working vs US/Australia where few employers will consider you, let alone make significant adjustments to allow you to work.
 

A.B.

Senior Member
Messages
3,780
Likes
23,077
I just realized how contradictory the psychobabbler position on receipt of sickness benefits is. They emphasize the importance of stress reduction. There is no doubt that being pushed to work even when sick, or having to fight for support while being sick, and existential fears are very stressful. The psychobabblers however insist that it's good to inflict this stress on patients.

There is certainly more than one way to see things.

PS: they also seem to have no problem recruiting the family to be less supportive towards the patient, which can also be quite stressful.
 

Esther12

Senior Member
Messages
13,774
Likes
28,358
It's also possible that most of the improvement came from those who had been ill for a shorter time at baseline, since illness duration at baseline correlated with fatigue at follow-up.
Yeah - I'd have liked more detail on this.
 

Simon

Senior Member
Messages
3,788
Likes
14,822
Location
Monmouth, UK
Doodling slightly here but have tried to present the results of this study in a clearer way.

Unemployment is in red while employment (full or part time) or being a student is in green. Onset is when they got ill, Contact 1 is typically 5 years later, contact 2 another six years on.

Clearly things have improved but Contact 2 remains a great deal worse than onset. It's possible that 11 years from onset (age 35 vs age 24 at onset), some people would not be working anyway due to raising families, but even so....

Oh, employment at onset wasn't split into full/part-time so interpret with caution.

upload_2014-12-9_7-19-28.png

Edited to correct age (thanks @Dolphin) and to give correcct percentages (previous graph was number of people with data, ex 92)
 

Attachments

Last edited:

Simon

Senior Member
Messages
3,788
Likes
14,822
Location
Monmouth, UK
Commentary: Improvement in CFS patients after Glandular Fever/Monoucleosis

I think this is an important study giving good data on and analysis of outcomes for patients who developed CFS after Infectious mononucleosis (or Glandular Fever, usually caused by EBV)
  • It's one of the largest long-term follow-ups I've seen
  • Has a relatively 'clean' cohort (CFS post-GF only)
  • Uses an objective measure, employment, as the primary outcome
  • The results are actually quite encouraging (though it was a young cohort, half of whom had been ill for only three years or less)
  • The authors recognise that Benefits can help return to work rather than arguing this is a secondary 'gain' of the illness
  • The results are also reported fairly:
The study conclusion said:
about half of younger patients with CFS with long-term incapacity for work got marked improvement including full or part-time employment.

[Self-management strategies, long-term sickness absence benefits providing a stable financial support, in addition to occupational interventions aimed at return to work were likely contributors to the generally positive, prolonged outcome]
The study comes from a Neurology department, in Haukeland hospital - of Rituximab fame. And although this wasn't a trial, patients were encouraged to use self-management, which sounded like pacing/activity management, and the authors concluded that this probably contributed to the relatively good outcomes:
A written self-management programme included information about the illness to provide the patients with a rationale and structural meaning for their illness experience.15 Active coping strategies for daily life included graded activity planning; encouraging activity, but staying within their physical limitations with consistent rest periods to minimise fluctuations in fatigue and symptoms
How the study worked (important)
Unfortunately, this wasn't a prospective study where you take a bunch of patients and follow them at consistent time points.

The first contact was any time between 1996–2006, the second and final contact (questionnaire follow-up) was in 2009, an average of 6.5 years later. There was huge variation within the sample, eg at the second and final contact in 2009 one patient had been ill for 24 years whereas the shortest illness duration was 5 years. So there is a huge amount of variation as to how long patients had been ill when they were followed-up.

At the first contact, patients had bean ill for a mean of 4.7 years, and again that hides a lot of variation. The median (midpoint) was only 3.2 years, spread into four 'quarters' like this:
upload_2014-12-11_13-45-28.png

So it's possible that patient who had only been ill for a short time are responsible for most of the improvement between first and second contacts. More on this later, but its seems like it isn't a massive factor.

Results

Pleasingly, the study used employment status as the primary outcome, which I graphed below. Unemployment is in red while employment (full or part time) or being a student is in green. Onset is when they got ill, Contact 1 is typically 5 years later, contact 2 another six years on (though there is a lot of variation in follow-up times, as I said earlier).

Clearly things have improved for many patients, but the overall situation at Contact 2 remains a great deal worse than onset.
upload_2014-12-11_13-52-17.png

Caution: It's possible that 11 years from onset (age 35 vs age 24 at onset) some people would not be working anyway due to raising families so unemployment might not be zero even for a healthy group. Please bear in mind that employment at onset wasn't split into full/part-time..

More on 'improvers'
The study measured 'improvement' in several ways. As well as change in employment status, they looked at
  • self-rated change (worse, stable, improved or recovered)
  • Fatigue, measured by the Fatigue Severity Scale, 1=no fatigue, 7=extremely fatigued; they then looked at CHANGE in fatigue between contact 1 and contact 2: a change of 1.0 or less was counted as stable (note the score is an average over 9 questions so a change of 1.1 is possible and would count as improvement/worsening)
The graph shows how patients improved on a number of different measures.
upload_2014-12-11_14-10-0.png

Note that half of those employed are now working full-time, compared with only 10% at contact 1, so presumably there has been an increase in hours worked per person, as well as more people working. [some of the gain may have been from full-time students at contact 1 becoming full-time employees at contact 2, we don't know if students were full or part time at contact 1].

One reason more people showed improvement in self-rated change than fatigue could be because they were measured in slightly different ways. Self-rated change includes any improvement, however small, while fatigue improvement required a mininum improvememt of 1.0 (0-6.0 possible for someone going from maximum fatigue to none).

Most people improved, even those who hadn't improved at Contact 1
Another interesting point was that of the 26 people who said they had alread improved at Contact 1, 25 improved again by Contact 2. And of the 38 who reported they hadn't improved before, 25 (66%) improved by Contact 2.

Fatigue improved much less than employment status

One slightly strange finding, which the authors didn't comment on, is that average fatigue levels falls rather modestly compared with the percentage improving in employment status. FSS scores were 6.4 at Contact 1, falling to 5.0 at Contact 2 - and 5.0 is the threshold set for 'severe fatigue' so this group of people, including those now working, are still pretty fatigued.

It seems likely that this in part is down to people getting back to work but still struggling, so that their level of fatigue doesn't improve as much as it might. Interesetingly, 13% rate themselves as 'recovered' compared with 28% who are working full time, which supports this view to some extent. On the other hand, Fatigue score was highly correlated with employment status (r=0.5, which is fairly impressive) i.e. those in employment were substantially less fatigued than those not in employment.

So perhaps what's happening is at least some people are improving and choosing to work at least part-time despite not being fully well.

Cognitive problems improved for a good number
The proportion reporting cognitive problems felll significantly between contacts 1 and 2 (both p<0.001):
Concentration problems: 83% => 58%
Memory problems: 72% => 51%
Many other symptoms reduced too, though to a lesser extent.

A minority got worse
  • 12% rated themselves worse overall
  • 8% showed deterioration in fatigue of 1.0 points or more
How much worse they got isn't clear. You would expect some people to get worse due to natural fluctuations (just as some will improve for the same reason even for individuals who are roughly on a level over time). Possibly some people ot catastrophically worse: that information wasn't reported.

The number of people getting a permanent Disability Pension, as opposed to merely sickness benefits, increased dramatically from 8 to 54, but it isn't clear if that was down simply to the longer of time they had been ill byy Contact 2, as opposed to any deterioration in health.

What 'predicts' return to work/improvement over time? (not a lot)
The authors ran some fancy models to see what predicted being in employment at final contact, or improved fatigue score.

For symptoms, less joint pain (arthralgia) did predict a better employment outcome but the p value was only 0.028 and almost certainly wouldn't have been significant if the results had been corrrected for mulitple comparisons. No other symptoms were significant predictors.

Looking at fatigue instead of employment, joint pain and depression and Contact 1, and eduction, were predictors - but between them these three factors only accounted for 22% of variance - out of a possible 100%, so there's nothing much going on here either.

It isn't clear if they looked at length of illness as a predictor [I will ask the study authors], so at least to me it still isn't clear to what extent the relatively short illness duration of many patients contributed to what appear to be better than expected results.
 
Last edited: