Reader Comments

Post a new comment on this article

Can the authors show the Sensitivity Analysis results for Societal Benefits for CBT & GET?

Posted by SMcGrath on 20 Aug 2012 at 15:26 GMT

McCrone and colleagues found that CBT and Graded exercise (GET) were cost effective for CFS on Healthcare costs alone. But they also emphasised that, from a societal perspective, the cost-effectiveness for CBT and GET relative to Specialised Medical care (SMC) was far greater. This was due to additional savings from the reduced need for informal care. They also state that from the societal perspective both CBT and GET are dominant to SMC, i.e give more benefit with lower costs, which is an ideal situation.

Missing Sensitivity Analysis Results
However, it appears that some of these conclusions would not hold if informal care was valued at the minimum wage of £5.93 rather than mean average earnings of £14.60 per hour. While this scenario was investigated in the sensitivity analysis, the actual results are not given. Instead there is a comment that changing the value of informal care did not have a 'large impact' on informal care costs, and that ‘the results were robust for alternative assumptions’.

The paper goes on to argue that investment in CBT & GET ‘would be justified in terms of improved quality of life of patients and “would actually be cost saving if all costs including societal costs are considered.” ’

Data from the paper suggests results would be substantially different if informal care was valued at the lower rate
The major contributor to CBT & GET Societal savings were from informal care, since lost employment was not significantly different between any of the groups.

Compared with SMC (and adjusted for baseline costs), CBT and GET showed informal care saving of £1,165 and £1,173 respectively when valued at mean average earnings of £14.60 per hour. Therefore, at the lower rate of £5.93, the minimum wage, these savings would reduce to £473 and £476 respectively.

The reduction in informal care costs would reduce societal gains, relative to SMC, by a corresponding amount:

Impact on societal costs (vs SMC) of using minimum wage to value informal care

CBT [informal care “@ mean earnings”]: -£698 (of which- £1,165 due to informal care]
CBT [informal care “@ minimum wage”]: -£6 (of which -£473 due to informal care)

GET [“mean earnings”]: -472 (-£1,173)
GET [“minimum wage”]: +225 (-£476)

This changes this situation substantially. GET now has a cost of £225 (as opposed to a saving of £472) while CBT is effectively neutral, with a per-patient gain of £6 gain at minimum wages instead of saving nearly £700. So the claim that: "there would actually be cost saving if all costs including societal costs are considered" appears not to hold in the minimum wage scenario (for GET and CBT combined), suggesting that the sensitivity analysis would show this finding was not robust. Similarly, GET would no longer be dominant to SMC as its societal costs would be higher; CBT would be technically dominant, but only thanks to a £6 margin per patients. Finally, the QALY-based cost-effectiveness acceptability curves (societal perspective, Figure 2) might look different in such a scenario.

My apologies if these calculations are flawed, and I trust the authors will point out any errors, but I only went down this route as the results of the sensitivity analysis were not available, when they could perhaps been provided as an appendix. Could the authors now publish the results of their sensitivity analysis for informal care?

No competing interests declared.

RE: Can the authors show the Sensitivity Analysis results for Societal Benefits for CBT & GET?

spjupmc replied to SMcGrath on 20 Aug 2012 at 21:29 GMT

Thanks very much for these helpful comments. It is encouraging that one of the key impacts of the interventions - the impact on informal care - has been addressed. The way that CFS/ME impacts on family members is crucial and it is surprising that no other comments have picked up on this. You are quite correct that valuing informal care at a lower rate will reduce the savings quite substantially, and could even result in higher societal costs for CBT and GET. The figures in the text relating to informal care are for all patients for whom we had cost data, while the the data in the table relate to patients for whom we had cost data and QALY data. I know this is a technical difference but it is important (as you can see from the text, the societal cost difference for CBT v SMC is quite a bit different from that in the Table due to this). I have looked at the data and valuing informal care at the minimum wage rate for those with cost and QALY data results in higher societal costs for GET of £3. As such GET does not dominate SMC at this level but rather has an cost per QALY of £87. This is of course somewhat below the NICE threshold of £30,000 but we are pleased to clarify this. It is of course arguable whether we should value informal care as low as this and in fact the minimum wage is now higher. What should be stressed above all else is that there is uncertainty around all of these cost and outcome estimates and therefore the acceptability curves are the more informative indicators of the relative cost-effectiveness of these interventions. Thanks again for the time you have taken to look at the data.

Competing interests declared: Lead author