Willingness to pay for medical treatments in chronic diseases: a multicountry survey of patients and physicians

A cross-sectional design, using a convenience sample, was used to identify the participants’ WTP. Patients eligibility criteria were chosen to be aligned with the indications of the products being studied for CHF and psoriasis. Inclusion criteria for psoriasis stakeholders were as follows: patients had to be at least 18 years of age and referred by a physician having been diagnosed with moderate to severe psoriasis; physicians had to be dermatologists with at least 5 years of postspecialization practice and regularly involved in the treatment of patients with moderate to severe psoriasis. Inclusion criteria for CHF stakeholders were as follows: patients had to be at least 40 years of age and referred by a physician has having been diagnosed with CHF with a reduced ejection fraction (HF-REF); physicians had to be cardiologists, internists or general practitioners with at least 5 years of postspecialization practice and regularly involved in the treatment of patients with CHF. For the CHF sample, physicians were divided equally between cardiologists and other practitioners (internists and general practitioners).

Exclusion criteria for all psoriasis and CHF stakeholders were a lack of proficiency in English or a local language and unwillingness or inability to provide an informed and signed consent.

For both the psoriasis and CHF samples, stakeholder groups were drawn from the public and private sectors so as to reflect the general distribution of public and private treatment patterns in the general population. For example, if, in a given country, 70% of the population was generally treated in the public sector and 30% was treated in the private sector, then the target sample for each stakeholder group comprised 70% of participants from the public sector and 30% from the private sector. With respect to sample selection by geographic area, the target sample for all stakeholder groups represented the two most populous cities in each country or the largest city and a city known to have the second-largest commercial center. Within each country, the number of stakeholders enrolled by city was proportional to the population in each city.

The selection of stakeholders for all groups was made using the existing Axios in-country networks, and the final groups were convenience samples that adhered as closely as possible to the criteria outlined above. Patients with moderate to severe psoriasis or CHF were identified through their treating physicians and approached for study participation immediately following a physician consultation (exit interview conducted on the premises of the institution where consultation took place). Physicians were identified through the Axios in-country networks and were initially approached for participation through an introductory letter, followed by a phone call. Physician interviews were conducted in their respective offices. The survey was conducted by Axios International in the framework of the establishment of a tracing strategy for two new products of a pharmaceutical company, using a slightly different selection of targeted countries depending on the product. The company had no role in the study design, analyses or drafting of the manuscript.

For both psoriasis and CHF medications, the patient questionnaire was organized based on several relevant domains identified from the literature as potential factors influencing WTP, including: socio-economic factors, in other words, household income, education/socio-economic status, insurance coverage (assessed through questions about the availability of any help from the government or an insurance company to pay for the treatment of the disease of interest and what percentage of the total cost was on average covered by this insurance); severity of the disease (assessed through the existence of lesions in sensible areas, namely the palms of hands, the soles of feet, the face or the genital region in patients with psoriasis, and the overall number of emergency room [ER] visits and the NYHA Class in patients with CHF) and current health status; and level of concern of disease health effects. The questionnaires were translated into the languages of the surveyed countries and pretested for comprehensibility and average duration of administration by staff speaking the local language and a selection of patients. The patient questionnaire was designed to solicit direct patient input on their own WTP for a new therapy, while the physician questionnaire captured physician perspectives on what they thought would be a reasonable price that their patients, not the physicians themselves, would pay.

The CVM has previously been described [10,11]. Briefly, the CVM was initially developed as a survey approach to assess the value of goods and services provided directly to the public without a consumer-based market and to determine how this valuation impacted price and demand for these goods and services. The approach has also shown utility in evaluating goods and services, including those related to health and medicine, for which a market exists but is regulated by the public sector and not based on free market economics. WTP is included as a component of both cost–effectiveness and cost-benefit analyses, and in the latter, can provide insight into how effectively goods and services are allocated.

Descriptive results are given as means ( ± standard deviation) for quantitative data and numbers (%) for categorical data. Unadjusted and multivariate analyses were conducted to identify independent determinants of the prices stated by patients and physicians. For unadjusted analyses, unpaired t-tests or Mann–Whitney tests (two groups) and ANOVA or Kruskal–Wallis tests (three and more groups) were performed to compare mean prices across categories, depending on the normality of variable distributions. Multivariate analysis was conducted using linear regression models led on all observations, since no patients declared a zero-valued WTP. A stepwise backward approach was used for sequentially removing variables not significant at the p < 0.05 level until reaching the final model.

Price values were assessed as international dollars using the World Bank’s gross domestic product purchasing power parity (PPP) conversion factor. A sensitivity analysis was also conducted based on nominal dollars to assess the robustness of the results. Similar results were yielded using this latter approach with respect to the independent predictors of patient’s WTP identified in multivariate analyses so that only results from the main analysis by $PPP are shown.

Finally, supportive descriptive results based on an in-house proprietary mathematical model (‘affordability model’) were computed in order to provide quantitative global numbers summarizing drug access and ability to pay of patients within a given country. Briefly, the model uses a three-step methodology, by determining the income distribution curve across population segments based on available data from international agencies (World Bank, IMF); estimating household healthcare expenditure, overall and specifically allocated for the treatment of interest; and finally limiting the patient pool based on all expected barriers that might impede the diagnosis, treatment and follow-up of patients apart from the ability to pay for medication, including ease of disease diagnosis, availability of sites to diagnose the disease and the rural rate. Results from the affordability model are shown as curves plotting the number or the proportion of diagnosed and treated patients expected to be able to access and buy a product against a range of hypothetical prices.

All analyses were stratified by chronic condition (psoriasis or CHF) and performed using Stata v14.1 (StataCorp, TX, USA).

The study was conducted in eight developing countries for CHF (Egypt, India, Indonesia, Mexico, Nigeria, Romania, Russia and Venezuela) and five developing countries for psoriasis (Egypt, India, Indonesia, Philippines and Thailand), and comprised a total of 167 CHF patients, 155 CHF physicians (102 cardiologists, 24 internists, 28 general practitioners and one other physician type), 85 psoriasis patients and 64 psoriasis physicians (dermatologists). One patient with CHF with too much missing information (including the primary end point) was excluded from the present analysis, yielding 166 CHF patients.

As summarized in Table 1, psoriasis and CHF patients had a mean patient age of 47 years and 62 years, respectively, with a higher education level in 51 and 40% of patients. Mean and median income values varied by country and by disease category. Overall, median monthly income (USD) and median monthly income ($PPP) were $560 and $1865 for psoriasis patients and $477 and $1031 for CHF patients. Median monthly income (USD) ranged from $210 (Egypt) to $1120 (Philippines) for psoriasis patients and $279 (Egypt) to $1272 (Venezuela) for CHF patients. Median monthly income ($PPP) ranged from $699 (Egypt) to $2801 (Venezuela) for psoriasis patients and $773 (Russia) to $3253 (India).

Figure 1 depicts the WTP for one CHF treatment and one psoriasis treatment according to education level (panels A and B), patients, physicians and affordability model (panels C and D) and country (panels E and F). The graphs represent the percentage of respondents who were willing to pay for a treatment (y-axis) according to the price for the treatment (x-axis). The curves revealed a clear association between higher WTP for expensive products and higher education level in both conditions, with an apparent more marked and graded relationship found in psoriasis patients (panels A and B). Patients’ WTP was found consistently higher than the cumulative percentages obtained from physicians and the mathematical ‘affordability’ model (panels C and D), indicating a gap between actual expectations and demands from patients and an external estimation of the price deemed reasonable by physicians or the patient’s financial capacity calculated by simulation. Finally, WTP varied strongly between countries and medications, with the highest WTP percentages found in Indonesia, Venezuela, Egypt and India for CHF and Thailand and Philippines for psoriasis (panels E and F).

This paper presents the results of a study on consumers’ willingness and ability to pay for new drugs developed for two different chronic conditions, CHF and psoriasis, in eight and five developing countries, respectively, using the CVM. The results from multivariate analysis indicate that country, recent angiotensin receptor blocker purchase and $PPP monthly income were significant factors impacting patients’ WTP for CHF treatment, while country, education level, monthly $PPP income and monthly $PPP out-of-pocket expenses for psoriasis treatment were significant factors for patients’ WTP for psoriasis treatment. For physicians, treatment sector was a significant factor in reasonable price for both CHF and psoriasis treatment, while country was a factor for CHF treatment and average treatment duration for patients currently treated was a factor in WTP for psoriasis treatment.

Previous studies have shown that stated inability to pay for treatment is a common reason for un-WTP for treatment [1]. In the current study, monthly $PPP income was a significant factor in patients’ WTP for both CHF and psoriasis treatment. This suggests that individual economic resources are a key determinant independent of the morbidity and mortality risk associated with a given disease. It is important to consider this economic factor when establishing pricing models, especially for diseases with higher long-term health risks that could ultimately results in higher patient care costs overtime. For example, while a patient who forgoes psoriasis treatment due to economic constraints is less likely to require more expensive care in the future, patients who forgo relatively cost-effective oral medication for CHF may require costly interventions or life-saving treatments in the future. Pricing models for diseases with significant morbidity and mortality should be designed to encourage early use of less costly therapies even among those patients with limited financial ability to pay in order to reduce or minimize the need for more costly care in the future [12,13].

In this study, the association between education level and WTP was significant only for psoriasis patients (persisting in multivariate analysis) but not for CHF patients, suggesting the influence of psychological factors related to social interactions and aesthetic concerns on WTP. This is consistent with a previous study demonstrating that dermatological diseases have a negative impact on health-related QoL even in the absence of serious physical limitations or discomfort [14]. Patients with higher education levels may be engaged in work-related activities in which their physical appearance may impact their ability to succeed or achieve career advancement, leading to an association between education level and a disease with aesthetic effects. Similarly, more educated patients may have higher aspirations for social status than their less-educated peers, increasing the value of personal appearance among the more educated patients. In contrast, no correlation was observed between education level and patient WTP for CHF treatment. This may be due to patients at all educational levels understanding the serious nature of CHF.

Price recommendations from physicians appear lower than WTP stated by patients across the factors assessed in this study. This discrepancy appears even more markedly in psoriasis patients, which is consistent with previous reports indicating that physicians tend to be more willing to prescribe more expensive therapies if they address diseases that are life-threatening rather than conditions that impact QoL but do not affect survival [15]. That patients with psoriasis are willing to pay more for treatment than their physicians recommend underscores a possible disconnect between the consumer and medical communities with respect to the negative effects that dermatologic diseases impose on patients’ day-to-day lives [14] and the related benefits they might expect from new therapy, to be compared with the physicians’ point of view who have a broad experience seeing how these treatments actually work in larger populations and real life settings.

Previous results suggest that WTP surveys are reliable and valid, and improve a program manager’s ability to predict client responses to price changes [11]. Our findings provide insight into factors that impact patients’ and physicians’ WTP for chronic disease therapies and may help guide the development of both pricing models and educational programs that align the actual and perceived physical and psychosocial benefits of different therapies. We identified patient financial ability to pay for therapy as a significant factor in WTP for both CHF and psoriasis treatment. From a price model perspective, models that provide additional subsidies for those patients with fewer financial resources may help increase use of cost-effective therapies earlier in the disease process, delaying or avoiding the need for most costly treatments in the future. This would be beneficial to both patients and countries’ health systems. For diseases with high risks of morbidity, mortality and increased costs of future care, educating patients about the benefits that specific treatments can provide may help increase the perceived value of these therapies and drive a similar increase in patient WTP.

The finding that educated psoriasis patients have a higher WTP for treatment than their physicians again suggests that there is a disconnect between how these two groups view the effects of the disease on daily life. Educational programs designed to help physicians understand the psychosocial impacts of psoriasis may be important for ensuring that they effectively address their patients’ needs.

Our study has several limitations. First, the sample size per country was limited and cross-country comparisons should thus be interpreted with caution, considering the multiplicity of factors possibly involved, including economic (gross domestic product), health insurance (out-of-pockets payments) and cultural elements and/or access to generic medications. Some potential relationships may also have been overlooked due to insufficient statistical power. However, despite the relatively low number of respondents, we were able to identify important factors that impact patients’ WTP and physician perception of a reasonable price to be paid by their patients for chronic disease treatments. Our findings also appear consistent with previous reports on the topic and are likely to be reproducible in other comparable settings and/or in larger surveys. Regarding the method used, we did not observe any ‘protest zeros’, which may occur when respondents reject some aspect of the contingent valuation (CV) market scenario by reporting a zero value even though they place a positive value on the amenity being valued, and thus may affect the reliability [16]. Second, considering the CVM used in the present study, a hypothetical bias cannot be ruled out, referring to the discrepancy between the hypothetical WTP stated by respondents in laboratory or field experiments and the actual (revealed) amount they would pay. It has been well documented in systematic reviews that hypothetical WTP values systematically overestimate the actual payments [17]. Despite this limitation which probably affected the accuracy of our estimates of absolute WTP values, it is noteworthy that the main objective of our work was focused on the identification of the factors associated with WTP, thus most likely preserving the relevance of our results. Finally, the study population was sampled from the two most populous cities from each country, thus precluding direct generalization to the whole population.