What can wage development before and after a G93.3 diagnosis tell us about prognoses for myalgic encephalomyelitis?, 2024, Kielland et al

[Nightsong]

Abstract:
Prognoses for persons affected by myalgic encephalomyelitis (ME) are rarely studied systematically. Existing studies are often based on smaller samples with unclear inclusion and subjective outcome criteria, and few look at wages as indicators of illness trajectories.

This article considers how ME affects the wages and dependency on public transfers of people affected over time, especially in the period when the welfare authorities investigate eligibility for disability pension.

We matched Norwegian population register data on 8485 working-age individuals diagnosed with G93.3 (postviral fatigue syndrome) from 2009 to 2018 with wage and transfer data and compared male and female cases to control groups.

The G93.3 population's wages fell sharply from around 3 years before diagnosis to 1 year after and stabilized at a low level. Public transfers started increasing several years before diagnosis and stabilized at a high level after. Few of those making no or very low income around the time of the diagnosis resumed earning moderate wages, and only exceptional cases returned to wages corresponding to median wages.

Link | PDF (Social Sciences & Humanities Open, open access)

 

[Nightsong]

One recommendation for/obligations of a person diagnosed with G93.3 under the AAP scheme is to attend a rehabilitation program. In this final section, we look at wages and taxable transfers in the years before and after attending rehabilitation programs. Kielland et al. (2023) reported that although many ME affected report to have enjoyed such programs for social reasons, they felt sicker afterward.

Fig. 6 shows wage and taxable transfer development before and after the rehabilitation programs. Some patients had been to two or more programs within the same calendar year, most often at the same rehabilitation center. To simplify the analysis, we operated with only the calendar year they had been to one or more such programs and compared the data to the years before and after. The graph shows how, on average, wages in all groups kept falling, and transfers kept increasing, in the year after the rehabilitation program (Y0).

These findings are significant for five reasons. First, going into treatment and rehabilitation interventions with the explicit goal of fully recovering labor capabilities may incite false hope in people often desperate to return to their former lives. Very few appear to recover fully, and the overall trend during the AAP intervention period is one of deteriorating capabilities. Second, health and welfare providers must address this group with realism. Long, demanding intervention periods for persons diagnosed with G93.3 to determine their disability status may discourage and push out the most vulnerable from accessing rights-based public welfare benefits.

Alternately, it could prolong their ordeal rather than yield returns in terms of persons recovering to more work capacity. Third, there is a potential trade-off between interventions aimed at improving life quality and those investing in regaining income earning capabilities for persons diagnosed with G93.3. Too strong of a focus on the latter may well be to the former’s detriment—and with minimal returns.

Fourth, the size of the population resembling ME patients is growing due to Long Covid, and this stresses the urgency of a more evidence-based approach to support the group. Lastly, absent an effective treatment for ME, more funding needs to go into research to find a cure. Whereas services and interventions for the group today are costly, they do not, in their current form, appear to have worked to fully rehabilitate more than the exceptional few.

 

[Kitty]

That's a really interesting approach.

I'm surprised there have been so few studies that looked at actual work or higher education outcomes in relation to rehab. Maybe policymakers aren't really that keen to know the answer.

 

[Denise]

This is interesting.
It would be nice if there were a way to take into account the change in wages of those who care for PwME....

 

[Simon M]

I've only just started reading this, but what an astonishing study. On the surface it is great methodology, covers the real world beyond clinical trials and focuses on objective outcomes. If it does what it says on the tin, it should prove very important in understanding the impact of the illness, provide powerful evidence for advocacy, and perhaps give us the best estimate yet for recovery/improvement rates.

I need to lie down, but here are a couple of snippets

Looks like a decent sample size (6-8.5k):

. Materials and methods
For the analysis, we used data from Norwegian patient registers (NPR) matched with income, transfer, and sick-leave data from Statistics Norway. The data covered registrations from 2009 through 2018 and included 73,357 observations of 8485 working-age (18–67 years) people in at least 1 year of the studied period. For the longer timelines, we used all observations of the 6249 individuals born between 1952 and 1991 and, therefore, of working age for the entire period studied.

Fig. 2. Comparing wage development for working age (18–67 years) men and women with a G93.3 diagnosis between the full sample (N = 6249) and the cohort diagnosed in 2016 (N = 1244).

 

[Kitty]

It's not un-useful to arguments about the gender pay gap either!

 

[Yann04]

Interesting (and a little depressing) study.

I’m actually suprised more people diagnosed with “postviral fatigue” didn’t fully recover. I couldn’t find anything specific in the article but it doesn’t sound like strict criteria.

Nice to see a critique of the rehabilitation model

going into treatment and rehabilitation interventions with the explicit goal of fully recovering labor capabilities may incite false hope in people often desperate to return to their former lives. Very few appear to recover fully, and the overall trend during the AAP intervention period is one of deteriorating capabilities.

Also really impressed by this:

Use of inclusive language
The target group for this article is often referred to with the term CFS/ME. We have consultet the APA Style Manual, chapter 7 on how to write with inclusivity and respect., and followed the advice to 1) seek guidance from our national self-advocacy group, and 2) asking participants we worked with directly through a survey (N = 660). We found that very few identified with the CFS-term and it was by the patient association explained as a trivializing way to describe a complext illness. We have therefore settled on using the ME term alone. We have, also with reference to APA, limited our use of the term “patient” to contexts of health-care settings.

 

[Mij]

@Nightsong

I love your new avatar

 

[Hutan]

I'm looking forward to reading this paper. I agree with others that this is a really interesting approach (also agree with the gender gap comment ).

I wonder how many countries would have similarly strong and linked databases of medical records and income information available to researchers? Possibly not many.

Also I agree, I love Nightsong's avatar picture. It's beautiful and uplifting.

 

[ME/CFS Skeptic]

Looks very interesting. Impressed by the work of Kielland so far.

Some take-away points:

- At the time of diagnosis most of the decline in income (and increase in sick leave benefits and transfers) already occurred. If you look at those years before diagnosis, the effect looks really strong compared to the control group (dashed line):


- Very few patients show significantly improvement in income suggesting the recovery rate is low. For example: "In the 2016 cohort, 57% of the 1524 cases showed no improvement, 38% deteriorated, and only 5% improved somehow over the 3 years studied."

- 415 G93.3 patients had data on following a rehabilitation program. Their wages kept falling one year after the program (they had been falling for several years before that) and stabilised afterwards. It would be interesting if they could compare this with the trajectory patients who did not follow rehabilitation. But as the authors suggest, it already indicates a very low succes rate. "going into treatment and rehabilitation interventions with the explicit goal of fully recovering labor capabilities may incite false hope in people often desperate to return to their former lives. [...] health and welfare providers must address this group with realism."

 

[Kalliope]

Oh great
A new publication from Tjenesten og MEg, a research project on ME patient's meeting with the health care system in Norway. They are doing a thorough job as researchers who actually get it.
Thread here about the project.

They have written a summary of this study:
Ny publikasjon fra prosjektet Tjenesten og Meg om utvikling av lønnsinntekt og offentlig stønader blant ME-syke

automatic translation:
New publication from the project The Service and ME on the development of wage income and public benefits among ME sufferers

quote:

The article concludes that the results are important for several reasons.

• NAV's ambition to restore ME sufferers' earning capacity through measures and treatment interventions can help to create unrealistic hopes among patients who often have a strong desire for recovery. Within the timeframes we have analyzed - which for parts of the group extends over 10 years - few seem to fully return to paid work, and the general trend is a negative development for most.
  
  
• Profits from existing measures in the form of improved earning capacity thus appear to be low.
  
  
• Annual incidence of G93.3 was increasing in the years for which we had data. This gives further reason to investigate how the public sector can best meet this group.
  
  
• In the absence of an effective treatment for ME, more should be invested in research to meet the patient group as best as possible. While services and interventions for the group are currently costly, in their current form they appear to have improved the earning capacity of only a few patients. The search for new and better treatment options should therefore be strengthened.

 

[Hutan]

Whereas services and interventions for the group today are costly, they do not, in their current form, appear to have worked to fully rehabilitate more than the exceptional few.

While services and interventions for the group are currently costly, in their current form they appear to have improved the earning capacity of only a few patients.

Just a little quibble - there seems to be a suggestion, especially in the summary article, that the services and interventions have worked for a small number of patients. We don't have any evidence of that - improvements could have been the result of cold-water bathing or stomach massage or whatever else the person was trying, or just some random change occurring during the passage of time.

As @ME/CFS Skeptic said, it would be interesting to compare the trajectory of those who undertook rehabilitation with those who didn't (although there will be a number of differences between the two groups that might muddy the waters).

 

[ME/CFS Skeptic]

It all depends however on how reliable those databases are.

- First there is the question of whether the diagnostic code G93.3 really reflects ME/CFS. In past database-studies of this kind we have seen abnormally high prevalence rates that increase in older age groups. Table 4 suggests that this cohort was younger with a mean age of 37 and 68% female predominance. Not sure what N= 42,889 and N= 30,468 refers to, probably the number of observations?

The authors do not report data of the entire cohort from which these ME/CFS were taken, so it's hard to estimate the prevalence rate. They note that Bakken et al. found 1162 cases per year in the same registries which corresponded to a incidence rate of 0.026%. In their less complete 2009- to-2015 sample, Kielland et al. found only 40% of the cases that Bakken found, so incidence was quite low. In short, it seems that the G93.3 diagnoses in Norwegian databases seem different that other registries. Perhaps Norwegian patients can comment on how reliable these registries are?

- Another problem is that there are a lot of incomplete cases, especially if you move further away from the diagnosis year. The trend for the 2016-2018 sample for which they had full data, however, looks the same as for the entire G93.3 sample. The authors also note that patients who fully recovered to no longer need medical care or had enough money to got to private clinics might not be recorded in NPR registries.

 

[Hutan]

Table 1shows that one in four cases had no wages the year before the diagnosis (2015), and this share gradually increased towards half the cohort 2 years after the diagnosis (2018). All categories of persons making more than NOK(2018) 100,000 also shrank, including those making more than NOK(2018) 500,000, from 13.8% to 5.7%.

That is a substantially lower participation in the workforce than I have seen suggested before. So, two years after diagnosis, 50% had no wages income, and only 5.7% were earning an income that the authors suggest is roughly indicative of a full-time income.

 

[Midnattsol]

I've been waiting on this study since I first heard Kielland mention it. Great that it is finally out!

The authors do not report data of the entire cohort from which these ME/CFS were taken, so it's hard to estimate the prevalence rate. They note that Bakken et al. found 1162 cases per year in the same registries which corresponded to a incidence rate of 0.026%. In their less complete 2009- to-2015 sample, Kielland et al. found only 40% of the cases that Bakken found, so incidence was quite low. In short, it seems that the G93.3 diagnoses in Norwegian databases seem different that other registries. Perhaps Norwegian patients can comment on how reliable these registries are?

Before the current guidelines for diagnosis in Norway that I think came in 2014, diagnosis could not be done by the patients GP in primary services. After the new guidelines it could, and primary services uses the diagnosis code A04 not G93.3 that is used in specialist health care.

I was diagnosed by my GP in 2016-2017, I have no idea if I am in this dataset.

 

[Midnattsol]

Also, something I don't think I saw mentioned when skimming through this: Some of us end up on student loans, which I think would make us disappear from this dataset.

Edit: Thinking about this a bit harder, we likely show up in the 0/low income group, so visible on income but no longer someone who receive benefits that is found in these registries*.

*Some people get education as a course of action when on AAP, however throughout the time period there has been changes in how this is done.

 

[RedFox]

The decline in income prior to diagnosis is stunning and proves what we knew all along: Many pwME have a delayed diagnosis, despite being very sick. It took me about 3 years 9 months.

 

[Midnattsol]

The decline in income prior to diagnosis is stunning and proves what we knew all along: Many pwME have a delayed diagnosis, despite being very sick. It took me about 3 years 9 months.

According to a survey by the Norwegian ME Association, the median time from symptom onset to diagnosis in Norway is ~5 years, the mean time is ~8 years.

 

[obeat]

@PhysiosforME a useful paper regarding rehabilitation outcomes.

 

[Sean]

A little disturbing that I feel happy about a paper confirming such appalling outcomes, including after 'rehabilitation' therapy.

But if it is robust and holds up then it is powerful ammunition on a number of fronts. Not least of all against the completely unproductive rehab approach, and the extraordinary cost of the disease to both patients and broader society.