Will (or can) people pay for headache care in a poor country?

The data contributing to this analysis came from a population-based survey of >1,000 respondents using sound and validated epidemiological methods [7]. Response rates were generally high: in the derivation of the original sample, household response rate was 95% overall, allowing little if any bias, whilst response rates within the sample for SF-36 (93%) and WTP (92%) were also very good. Only for MIDAS (70.5%) was the rate less satisfactory. It may not be compatible with Georgian culture voluntarily to admit that headache causes lost time, or it may be that those who did not lose time saw the questions as pointless (these possible explanations being counter-balancing in whom they would exclude, again making significant bias unlikely). WTP data were collected contemporaneously with other data allowing diagnosis and describing headache-attributed burden. This made possible comparisons between diagnoses, and correlations between WTP and burden.

We diagnosed episodic headache according to ICHD-II [9] as far as was possible, although 102 cases (17.8% of episodic headache cases) were unclassifiable (and excluded from these analyses). Headache occurring on ≥15 days/month might be any of a group of disorders, including chronic MIG, chronic TTH and medication-overuse headache (MOH), that are often but imprecisely referred to collectively as “chronic daily headache”; we preferred to avoid this term. Headache on ≥15 days/month proved very difficult to diagnose by questionnaire, and might in some cases require multiple diagnoses; therefore we analysed all such cases together, as a separate group.

Headache-attributed burden is multidimensional, and extends beyond the person immediately affected to family, friends and work colleagues, and to society as a whole. It is impossible to estimate it in its entirety. We have considered aspects that appear of particular importance—HRQoL and lost productive time. WTP is also an indicator of burden.

SF-36 demonstrated that HRQoL was reduced by headache but this was not quantitatively dependent upon headache type or frequency. The failure of SF-36 to differentiate between episodic headache and headache on ≥15 days/month is particularly surprising, but here is not the place to speculate in detail on the cause. MIDAS, on the other hand, revealed a greater burden from headache occurring on ≥15 days/month than from episodic headache, which is expected.

HRQoL reflects people’s assessment of their general well-being and position in life as perceived within the context of their culture, value systems, goals and concerns [15]. SF-36 is the most widely used HRQoL questionnaire in patients with chronic diseases [11, 15] and has been utilized in several epidemiological and clinic-based studies on headache, all agreeing that people with MIG have lower SF-36 scores than population controls [2, 17]; in one, people with moderate to high disability from MIG had lower HRQoL scores in all SF-36 domains [2] suggesting a profound impact on HRQoL. Other studies have looked at other headache types in addition to MIG [18‐20], always finding HRQoL to be negatively affected.

The MIDAS Questionnaire, on the other hand, reflects disability. Originally designed as a screening instrument for people severely affected by headache, who might most benefit from medical care, and to provide an outcome measure for clinical practice [21‐26], clinical trials [24, 27‐31] and epidemiologic studies [10, 18, 32‐39], it is a measure of behavioural response to impairment rather than of disability itself, producing scores expressing lost useful time. Like HRQoL measures, it is intended to aggregate the impact of illness on an individual over a period of time; unlike SF-36, it is disease specific, although applicable to headache rather than only to MIG [40]. It is essentially sensitive to frequency, which may explain why, in our study, MIDAS detected a relationship between headache frequency and burden while SF36 did not. Being disease specific, MIDAS cannot be applied to non-headache controls, who, of course, lose no time from headache. In other words, attribution to headache is explicit in the case of MIDAS, but implicit and inferred from comparisons with non-headache controls in the case of SF-36.

Together, these measures and the prevalence data reveal a population burdened by headache, with unmet need for health care. This is the context in which we explored WTP.

The results of the WTP enquiry are of interest for several reasons. First, we found that >90% of people with headache were willing to pay, out of their pockets from their generally very limited resources, for effective headache care. This is a striking argument against the view that headache is a problem of wealthy countries and unimportant in low-income countries such as Georgia. Second, while WTP varied with financial means (results not shown) as might be expected, it did so within a relatively narrow range (75% of both wealthier and poorer groups would pay between USD 5 and USD 10). This suggests that WTP is driven strongly by need, and only within limits by ability to pay, which is an important finding. However, our method of assessing wealth was inexact, because income in Georgian families is frequently hidden, or provided by a family member who lives and works outside the country and therefore not registered. Direct questions about income are not welcome, and responses would not necessarily reflect reality. Furthermore, in the rapid transition to a US-style market economy, many people with university education lost their jobs and became poor, while others, mostly young people, have prospered better even without education. Therefore, normal socioeconomic indicators do not work well, and so this was an imprecise analysis. Third, WTP did not depend on headache type or frequency. This demonstrates that people with less-frequent headaches are similarly interested in treating them to those with near-daily headache, and strongly suggests that a headache service, if available in the country, would be used not only by the minority of people with headache on ≥15 days/month but rather by the entire population of headache sufferers, seeking to reduce their personal burden of headache and to have a better quality of life. Fourth, of course all depends on alignment between WTP and the actual cost of the service. Whilst a WTP of USD 8 per month might seem rather low compared with what is often spent in western countries, it would be sufficient for many people when set against the low general costs in Georgia. Fifth, since not only the wealthy but also people with low income were willing to pay at least something, we believe these results are very important for market analysis by the pharmaceutical industry, who should have an interest in introducing modern anti-migraine drugs (e.g., triptans) to the Georgian market. The importance of this is in the following: without effective drugs, headache services will remain limited, and, while headache services are limited, the market for effective drugs will remain depressed. This Catch-22 situation needs urgently to be breached.

There is one caveat: we have recorded what people say they would do in the (currently) abstract circumstances of available good care; it is not certain, until empirically tested, which we hope later to do, that they will actually do it when the opportunity becomes real. Regardless of this, and a sixth point of interest, WTP is an expression of burden that probably captures more elements of it than either HRQoL or disability measures.

We have something more to say about headache care in Georgia. During the survey, we received the impression that headache was not considered, by those affected, an important medical problem. The majority of respondents considered their headaches, whether MIG or TTH, to be a natural part of their lives; people were surprised to learn that headache could be treated effectively. Many were very enthusiastic to hear that efforts were being made by Lifting The Burden [4, 5] to establish a headache service in the country. Funding of medical care in Georgia is a major issue. In the Soviet era, the State guaranteed the necessary minimum of wealth, and basic health-care services were provided with no out-of-pocket payments. Transition from this socialist system to a market economy has been accompanied by a marked socioeconomic decline for many inhabitants, and the development of significant disparities. Most people are not insured, and all costs for any treatment are out-of-pocket expenses.

In these circumstances, would people, in a poor country, pay for a service for primary headache disorders, which are not life-threatening but “only” reduce quality of life? We have found the answer to be “yes”, with the caveat referred to earlier. We assume that the answer would be the same in other similarly poor countries, and believe that efforts to introduce headache services in such countries are far from futile, and must be continued.