Impoverishment due to out-of-pocket health care expenses in Peruvian households

1 Introduction

In 2015, the member states of the United Nations adopted the Sustainable Development Goals (SDGs) with targets for 2030, including the eradication of poverty and the improvement of the health and well-being of the population (United Nations, 2015). A central component of this agenda is universal health coverage, aimed at ensuring access to necessary health services without households facing financial hardship. However, one of the main obstacles to achieving this goal remains out-of-pocket health expenditure, defined as direct payments made by households when receiving medical consultations, medications, hospitalizations, or other health services, financed with their own resources (Brañas-Garza et al., 2009; OIT & OPS, 1999; World Health Organization, 2010).

In this context, Peru continues to have relatively high levels of out-of-pocket health spending within Latin America, reflecting persistent limitations in household financial protection (Table 1). According to the World Health Organization, in 2020 these payments accounted for 22.8% of total current health expenditure, exceeding the average for high-income countries (17.1%) and the levels observed in the United States (9.9%), Colombia (13.6%), and Uruguay (16.7%). However, Peru's level remained below that of Mexico (38.8%), Chile (29.4%), and Argentina (24.2%). At the regional level, Peru's levels were lower than those in Africa, the Eastern Mediterranean, and Southeast Asia, though higher than those in the Western Pacific, suggesting partial progress in reducing out-of-pocket household expenditures (WHO, 2023).

Table 1. Out-of-pocket health expenditure (percentage), 2000–2020.

Prepared using data from the World Health Organization (WHO).

AFR, African Region; AMR, Region of the Americas; EMR, Eastern Mediterranean Region; EUR, European Region; SEAR, South-East Asia Region; WPR, Western Pacific Region.

A high reliance on out-of-pocket spending can have significant adverse effects on households' economic well-being (Akazili et al., 2017; Ebaidalla and Ali, 2019; Hooda, 2017; Yap et al., 2018). Part of this situation stems from limitations in the public provision of health services, such as low-quality care, drug shortages, long waiting lists, or insufficient coverage, which encourages greater use of private services. Furthermore, recent evidence shows that households with some form of health insurance continue to incur out-of-pocket payments due to coverage exclusions or unfunded ancillary costs (Alvi et al., 2020; Das et al., 2020; Ku et al., 2019; Wagner et al., 2018). Since these out-of-pocket expenses are often unexpected and unavoidable, many families resort to savings, borrowing, or selling productive assets to finance medical care. In more severe situations, households may reduce consumption of essential goods or postpone treatments due to a lack of resources. From an economic perspective, health financing based on out-of-pocket payments is considered one of the most inequitable and inefficient mechanisms, as it shifts the financial risk of illness to households and disproportionately affects lower-income groups (Dobridge, 2018).

When this financial burden reaches excessive levels, out-of-pocket spending may lead to catastrophic health expenditure and, subsequently, to impoverishment (Garg and Karan, 2009; Hooda, 2017; Keane and Thakur, 2018; Khan et al., 2017; Lavilla, 2012; Van Doorslaer, 2007). According to the Pan American Health Organization (2014), health expenditure is considered catastrophic when out-of-pocket payments account for 30% or more of household disposable income, although this threshold may vary depending on the context and country (Lavilla, 2012). Beyond this, impoverishment occurs when health payments reduce household resources below the real subsistence expenditure level (Xu, 2005).

Despite the relevance of this issue, empirical evidence for Peru remains scarce, particularly regarding recent trends and the socioeconomic determinants of impoverishment associated with health spending. Likewise, the COVID-19 pandemic may have altered both the magnitude and the distribution of household financial vulnerability. In this regard, the present study has two objectives: first, to estimate the extent of impoverishment associated with out-of-pocket health expenditures incurred by Peruvian households between 2010 and 2021; and second, to identify associations with the socioeconomic, demographic, and environmental factors linked to such impoverishment. In addition, the study analyzes the relationship between the COVID-19 pandemic and the magnitude and distribution of impoverishment related to health care expenditures borne directly by users.

The remainder of the article is organized as follows. Section 1 presents the introduction and literature review. Section 2 describes the empirical strategy and data used in the research. Section 3 presents and analyzes the results. Section 4 discusses the findings, including policy recommendations, limitations, and future lines of research. Finally, Section 5 presents the conclusions.

1.1 Literature review

The literature on health-related out-of-pocket impoverishment has established evidence that health systems heavily reliant on direct household financing tend to weaken financial protection, exacerbating preexisting inequalities and creating new situations of poverty. When such out-of-pocket expenditures a significant proportion of disposable income, they constitute catastrophic expenditures, which can force a reduction in essential consumption, in debtedness, or the sale of productive assets.

International evidence suggests that this problem is of considerable magnitude and varies across regions. In Asia, Van Doorslaer (2007) study, based on fourteen countries that account for the majority of the regional population, estimated that 2.7% of the population fell below the US$1-per-day poverty line after making out-of-pocket health payments. This finding demonstrates that health care spending can act as an immediate mechanism for widespread impoverishment. Similarly, India repeatedly emerges as one of the most critical cases: Hooda (2017) estimated that 3.5% of the population—equivalent to 50.6 million people—fell into poverty due to this cause, while Garg and Karan (2009) identified an additional 3.2% increase in poverty, with greater severity in rural areas. Subsequently, Keane and Thakur (2018) calculated that 4.1% of the Indian population (50 million people) remained in medical-expense-induced poverty. In Bangladesh, it was estimated that 3.5% of the total population falls into poverty each year due to out-of-pocket healthcare spending (Khan et al., 2017). In India and Bhutan have also highlighted the link between catastrophic health spending and impoverishment (Sharma et al., 2023). In Indonesia, out-of-pocket health expenditures contribute to the incidence and intensity of catastrophic and impoverishing health expenditures among households (Aji et al., 2013).

In sub-Saharan Africa, where average income levels are lower, the relative impact tends to be even more pronounced. In Ghana, Akazili et al. (2017) demonstrated that out-of-pocket payments increased poverty by 9.4% below the $1.25-per-day line and by 3.8% below the $2.50 line, showing that the effect varies depending on the monetary threshold used. In Sudan, Ebaidalla and Ali (2019) found that, after accounting for health care spending, the number of poor households increased by 10% nationwide, with a 17.4% increase in urban areas and an 8.7% increase in rural areas. This difference is significant because it challenges the traditional notion that health vulnerability is concentrated exclusively in rural areas; in certain urban contexts, higher monetary costs of care and reliance on private services can lead to more severe financial shocks.

Furthermore, the literature suggests that impoverishment does not always result from general use of the health care system, but rather from specific episodes of care that involve high direct and indirect costs. In Uganda, Yap et al. (2018) reported that nearly one-third of households incurred catastrophic expenditures associated with pediatric surgeries, even though partial free-of-charge schemes were in place. In India, Das et al. (2020) found that childbirth care represented a loss equivalent to 35% of wages, affecting poor women most severely. Similarly, Alvi et al. (2020) found that two-fifths of patients undergoing antiretroviral therapy fell below the poverty line.

In Latin America, although the region has higher levels of healthcare institutionalization compared to other developing economies, out-of-pocket spending continues to place significant strain on households. In Chile, Correa-Burrows (2012) estimated that families spent 5.4% of their total income on healthcare, a significant proportion of the household budget. In Peru, Lavilla (2012) found that 1% of non-poor households fell into poverty as a result of out-of-pocket medical expenses. Although this figure appears lower than those observed in Asia or Africa, it is significant because it reflects that even systems with greater state involvement fail to completely eliminate financial risk, especially for households near the poverty line or exposed to recurrent illnesses.

In light of this situation, the literature has also identified successful mitigation efforts through insurance policies. In the Philippines, Wagner et al. (2018) found that expanding health insurance reduced out-of-pocket spending by 21%, while in Taiwan, Ku et al. (2019) demonstrated that the National Health Insurance reduced the share of health care spending in total household expenditure by 2.11 percentage points. Both cases show that prepayment systems and universal coverage can significantly reduce financial vulnerability. However, comparative evidence suggests that these results depend on the actual scope of coverage, the availability of medications, and the quality of services, meaning that formal enrollment does not always translate into effective protection.

The determinants of impoverishment due to out-of-pocket health expenditure can be classified into several microeconomic groups, including socioeconomic, demographic and environmental factors. Among the socioeconomic and demographic factors that contribute to impoverishment due to out-of-pocket health expenditure are education, income, household expenditure, number of household members, age, sex, minor children, presence of women of childbearing age, being an adult in the household, ethnicity, health insurance coverage, etc (Correa-Burrows, 2012; Ebaidalla and Ali, 2019; Petrera Pavone and Jiménez Sánchez, 2018; Woldemichael et al., 2016). It has even been shown that in countries with universal health coverage, individuals may incur out-of-pocket health expenses, with impoverishing consequences (Alvi et al., 2020; Das et al., 2020; Ku et al., 2019; Wagner et al., 2018). In relation to environmental factors that affect impoverishment through out-of-pocket spending on health care, access to drinking water, sewerage and electricity, and housing conditions can be highlighted (Lavilla, 2012; Petrera Pavone and Jiménez Sánchez, 2018).

2 Materials and methods

2.1 Data and sample

We used microdata from the National Household Survey (ENAHO) at the national level with annual cuts between 2010 and 2021, which were combined into a pool. This survey, conducted by the National Institute of Statistics and Informatics (INEI), utilizes a probabilistic sampling of areas, stratified, multistage, and independent within the territorial scope of study, with a confidence level of 95% for sample estimates. The resulting information is publicly accessible and allows exploration of the standard of living of the population.

For this study, we focused on the sample provided by ENAHO exclusively for household heads aged 14 and older, for several reasons. Firstly, household heads tend to be the main decision-makers regarding the economic and financial affairs of the household, thus becoming an essential source of information on family socio-economic conditions. Furthermore, this focus ensures greater consistency and comparability of data over time, as household heads provide comprehensive information on income, expenses, and other relevant aspects of the household. The sample size selected was 21,496 household heads in 2010 and 34,245 in 2021, resulting in a total of 376,016 observations for the last 12 years.

2.2 Description of the variables

The dependent variable of the study was impoverishment due to out-of-pocket healthcare expenses, defined as the situation in which a household falls into poverty as a result of direct payments for medical services, excluding those related to transportation and special nutrition (Xu, 2005). This indicator was developed in two stages:

2.2.1 Calculation of out-of-pocket health expenditures

In the first phase, out-of-pocket health expenditures were calculated using the methodology of the World Health Organization (Xu, 2005) and previous empirical studies conducted by the General Directorate of Medicines, Supplies, and Drugs of Peru (DIGEMED, 2021) and Lavilla (2012). The information on out-of-pocket health expenditures was constructed with ENAHO and is available in three frequencies: The first group is on a weekly frequency (last 4 weeks), corresponding to expenditures on consultations, medicines/supplies, analyses, x-rays, tomography, and other exams. The second group is on a monthly frequency (last 3 months), corresponding to dental and related services, ophthalmology services, purchase of glasses, vaccinations, child health check-ups, contraceptives, and other expenses (orthopedics, thermometers, etc.). Finally, the third group is on an annual frequency (last 12 months), corresponding to expenditures on hospitalization, surgical interventions, pregnancy check-ups, and childbirth care. The data on out-of-pocket health expenditures are identified in ENAHO as variables I41601 to I41616, which are imputed, deflated, and annualized with prices from Metropolitan Lima (base inflation 2021) to obtain the real expenditure.

2.2.2 Methodology for identifying the relationship between factors contributing to impoverishment due to out-of-pocket healthcare expenses

In the second phase, once out-of-pocket health expenditures had been determined, the impoverishment associated with those expenditures was calculated using the World Health Organization's methodology (Xu, 2005).Where yh represents impoverishment due to out-of-pocket health care expenses; yh takes the value of 1 when the household was not initially poor, but after deducting health care expenses, its expenditure falls below the subsistence threshold; conversely, yh takes the value of 0 when the household remains above the subsistence threshold after such deduction; exph is the household's total real per capita expenditure on goods and services; seh is the real subsistence expenditure, calculated as the product of the monetary poverty threshold and the number of household members; and opph is the out-of-pocket health expenditure calculated in the first phase.

yh=1ifexph≥sehyexph−opph<seh

yh=0ifexph≥sehyexph−opph≥seh

Regarding the relationships with the independent variables, the economic literature groups them into various microeconomic categories. In this study, these variables were classified into three main groups: socioeconomic, demographic, and environmental factors (Table 2).

Table 2. Variables used in the estimation.

Socioeconomic factors include indicators related to education, income, and health conditions within the household. The total number of years of schooling completed by household heads is considered, as well as the natural logarithm of the total per capita income. Additionally, the number of household members generating income and the ratio of economically dependent people to income earners in the household are included. Dichotomous indicators reflecting the presence of chronic discomfort and permanent disability in the household, attendance at health centers, and affiliation with the EsSalud health insurance are also analyzed.

Demographic variables focus on the demographic characteristics of the head of the household and its members, both as discrete and dichotomous variables. The age of the head of the household and its square to capture nonlinear effects are considered. Additionally, the sex of the head of the household, the total number of people in the household, and the presence of children in different age ranges (under 3 years, 3 to 12 years, and 12 to 17 years) are included. The presence of pregnant women in the household is also analyzed using a dichotomous indicator.

Finally, environmental variables address the living conditions and the environment in which the household is located. Dichotomous indicators are used to identify households in inadequate housing (Unmet Basic Needs 1 - UBN 1) and those in overcrowded conditions (Unmet Basic Needs 2 - UBN 2). Additionally, it is considered whether the dwelling lacks adequate sanitary services and the type of fuel used for cooking, specifying the use of coal, firewood, or dung.

2.3 Estimation techniques: logit model

To estimate the associations between impoverishment due to out-of-pocket health expenditures and the determining variables, a nonlinear logistic regression model was employed. The logit model can be derived from a latent model, which is not observable but is related to the dependent variable (Wooldridge, 2012). In this case, the probability that an individual becomes impoverished or not by out-of-pocket spending depends on a critical threshold.

Let us assume a critical threshold yhuc, if yh∗ exceeds yhuc, then the household becomes impoverished. It is important to note that both yhuc and yh∗ are not observable. However, if uh is assumed to follow a logistic distribution, it is possible to estimate the parameters of yh∗.Where subscript h represents household heads, subscript r represents the departments of Peru (24 departments), and subscript t represents the years of study (12 years); yh is the probability of impoverishment due to out-of-pocket health expenditure; Sh is the vector of socioeconomic variables Dh is the vector of demographic variables Ah is the vector of environmental variables Rr y λt are the departmental fixed effects and the time fixed effects with pivot in the year 2019. The parameters θ, ϕ and ν show the impacts of the probability of socioeconomic, demographic and environmental factors on the probability of impoverishment due to out-of-pocket health spending. Likewise, the parameters associated with λt —specifically those for the years 2020 and 2021—would indicate the probability that COVID-19 would lead to impoverishment due to out-of-pocket health care expenses.

yh∗=φ+θSh+ϕDh+νAh+αRr+δλt+uh=xhβ+uh

yh=1siyh∗>0

yh=0siyh∗≤0

Thus, the probability of impoverishment due to out-of-pocket expenditure yh scan be represented as:Where F(xhβ) is a logistic cumulative distribution function.

Pr(yh=1|xh)=Pr(yh∗>0|xh)=Pr[ui>−xhβ|xi]

=1−F(−xiβ)=F(xhβ)

F(xiβ)=eφ+θSh+ϕDh+νAh+αRr+δλt1+eφ+θSh+ϕDh+νAh+αRr+δλt

The model is estimated using the maximum likelihood method,The first-order conditions of the maximization problem require that:One of the most widely used methods is the so-called iterative “scoring” method. The scoring method uses information from the information matrix (Maddala, 1983): βf+1=βf+I^−1(βf)∂β∂βf. The assumptions of logistic regression require that the errors are independent.

ln⁡L(β)=∑i=1n⁡{yhln⁡F(xhβ)+(1−yh)ln⁡(1−ln⁡F(xhβ))}

∂ln⁡L(β)∂β=∑i=1n⁡[yhf(xhβ)F(xhβ)+(1−yh)−f(xhβ)1−F(xhβ)]xh

3 Results

3.1 Structure of out-of-pocket health expenditures and impoverishment

Peru, a middle-income country, has historically recorded a high share of out-of-pocket health spending (Table 1). However, over the last two decades, out-of-pocket payments declined from 40.3% of total current health expenditure in 2000 to 22.8% in 2021 (WHO, 2023).

The distribution of out-of-pocket health care spending in Peru has remained stable over the last 12 years, with the purchase of medicines being the main item, representing an average of 38.7% (Table 3). The second most important item is dental and convex services, with an average of 16.1%, but this expenditure was reduced during the COVID-19 pandemic to 12.9% and 10.9% in 2020 and 2021, respectively. On the contrary, other expenditures (such as orthopedics and thermometers) increased significantly during the pandemic period to 22.7% and 28.1, by 2020 and 2021, respectively.

Table 3. Evolution of the distribution of out-of-pocket health care spending in Peruvian households from 2010 to 2021 (in percentages).

The results were normalized using the population expansion factor. INEI. ENAHO 2010–2021.

The place of healthcare provision in Peruvian households remained relatively stable between 2010 and 2021, although there were significant variations during the COVID-19 pandemic (Table 4). On average, drugstores and pharmacies were the main place of care, with a growing share over the period: they rose from 19.5% in 2010 to 24.7% in 2019, and although they declined slightly during the pandemic (24.3% in 2020 and 22.7% in 2021), they continued to be the most widely used option for households. This trend suggests that pharmaceutical establishments have become the most accessible and frequent mechanism for resolving health problems, especially in contexts of mobility restrictions.

Table 4. Evolution of the place of health care in Peruvian households from 2010 to 2021 (in percent).

The results were normalized using the population expansion factor. INEI. ENAHO 2010–2021.

Public facilities also played an important role, although with a smaller share. MINSA Health Posts recorded an average of 7%, while EsSalud hospitals accounted for around 4.6% of total consultations during the 2010–2021 period. However, both services experienced a significant decline during the pandemic, especially MINSA Health Posts, which dropped from 7.3% in 2019 to 5.1% in 2020. This decrease coincides with the implementation of health emergency and social distancing measures—such as the National State of Emergency and the lockdown established through Supreme Decree No. 044-2020-PCM—which restricted regular access to health services.

Additionally, the analysis shows that, on average, 49% of households did not seek care from any type of facility during the period analyzed. This proportion increased substantially during the pandemic, reaching 54.9% in 2020 and 55.6% in 2021, revealing a significant setback in access to health services. The increase of more than 9 percentage points relative to the previous trend suggests a combination of factors: fear of contagion, mobility restrictions, reduced availability of non-essential services, and the prioritization of COVID-19 care in public facilities.

With respect to impoverishment due to out-of-pocket health spending, and following the methodology proposed by Xu (2005), a progressive increase in this indicator is observed over the last twelve years. In 2010, 2.0% of non-poor households fell into poverty due to direct health expenses; by 2021, this proportion had risen to 3.6%. During the COVID-19 pandemic, the phenomenon reached its highest level, climbing to 4.6%, meaning that nearly 5 out of every 100 non-poor households became poor as a direct consequence of medical care expenses. This impact was particularly severe among rural households, male-headed households, and those located in the rural highlands and rural jungle. Furthermore, when considering the type of health insurance, households affiliated with the Seguro Integral de Salud (SIS) exhibited higher levels of impoverishment compared to those insured by EsSalud (Figure 1), reflecting the lower financial protection offered by SIS and the greater economic vulnerability of its affiliates to direct health expenditures.

3.2 Factors associated with impoverishment due to out-of-pocket health care expenses in Peruvian households

Table 5 presents the results of the estimated logit model used to identify the associations with impoverishment due to out-of-pocket health spending, based on a pooled dataset covering a twelve-year period. The model incorporates time fixed effects and departmental fixed effects, which allows for controlling both macroeconomic shocks and institutional reforms as well as structural characteristics specific to each department. The global statistics of the model indicate solid performance: the likelihood ratio (LR) tests are highly significant (p < 0.001) in all three specifications (total, urban, and rural), indicating that the included variables jointly explain the probability of falling into poverty due to health expenditures. The pseudo R-squared values (0.12, 0.16, and 0.08) are consistent with logit models based on household microdata and low-incidence events.

Table 5. Logit model estimation results.

Individual significance: * p < .05; ** p < .01; *** p < .001. Coef. refers to the estimated coefficient of the logit model.

The results show that impoverishment due to out-of-pocket health spending is associated with a broad set of factors operating across socioeconomic (Sₕ), demographic (Dₕ), and environmental (Aₕ) dimensions. In the socioeconomic dimension, variables such as education level, household income, the number of income earners, and the economic dependency ratio are linked to households' capacity to cope with unexpected health expenses. Lower schooling, lower income, and higher dependency ratios are associated with greater vulnerability to impoverishment, while the presence of illness, disability, or discomfort is associated with higher demand for medical services and, therefore, greater costs.

Demographic factors related to the life cycle and family structure are also associated with the probability of falling into poverty for health-related reasons. Age, sex, household size, and the presence of young children or pregnancies are linked to greater care needs, which may raise out-of-pocket spending, especially in households with vulnerable members or in advanced stages of life, when chronic diseases are more prevalent.

Finally, environmental factors reflect structural conditions associated with greater exposure to health risks as well as a higher likelihood of incurring substantial expenses. Overcrowding, lack of adequate sanitation, and the use of polluting fuels for cooking are associated with a higher incidence of respiratory and gastrointestinal diseases, generating greater demand for medical care and widening inequality gaps that translate into economic vulnerability.

Based on the time fixed effects—using 2019 as the reference year—it is observed that, in 2020, the probability of impoverishment due to out-of-pocket health spending was significantly higher than in the reference period. This pattern was particularly notable in urban areas, where the probability increased in a statistically significant manner, whereas no statistically significant change was observed in rural areas. These findings suggest heterogeneous changes across areas during the COVID-19 pandemic period, with urban households showing greater exposure to high health expenditures during the first year of the health emergency.

Subsequently, marginal effects evaluated at the mean were estimated in order to interpret changes in the probability of impoverishment associated with and related to marginal variations in each explanatory variable. According to the results in Table 6, the socioeconomic factors significantly associated with and related to a lower probability of impoverishment due to out-of-pocket health spending are years of education, per capita household income, attendance at health facilities, and coverage by EsSalud health insurance. Specifically, one additional year of education is associated with a 0.05 percentage-point reduction in the probability of impoverishment, while a 1% increase in per capita household income is related to a 0.18 percentage-point reduction. Likewise, attending a health facility and being covered by EsSalud are associated with reductions of 0.09 and 0.5 percentage points, respectively.

Table 6. Marginal effects of the logit model (at mean).

Individual significance: * p < .05; ** p < .01; *** p < .001; dy/dx are marginal effects. The model includes departmental and temporary fixed effects to control for regional differences and time-specific shocks.

Conversely, other socioeconomic factors were positively associated with the probability of impoverishment. The number of income earners, the economic dependency ratio, the presence of chronic discomfort, and permanent disability in the household were associated with increases of 0.19, 1.91, 0.34, and 0.30 percentage points, respectively, in the probability of impoverishment. These associations were significant in both urban and rural areas, indicating that the economic and health constraints faced by households were similarly related to impoverishment across the territory.

Regarding demographic factors, variables such as the sex of the household head, family burden, children's ages, and the presence of pregnancy were associated with changes in the probability of impoverishment. In particular, households headed by men exhibited a probability of impoverishment that was 0.4 percentage points higher than those headed by women. Each additional household member was associated with a 0.2 percentage-point increase in the probability. Children under 3 years of age and those between 3 and 12 years were associated with increases of 0.064 and 0.029 percentage points, respectively, whereas adolescents between 12 and 16 years were associated with a 0.19 percentage-point reduction, possibly reflecting lower requirements for intensive medical care. Likewise, the presence of a pregnancy in the household was associated with a 0.98 percentage-point increase in the probability of impoverishment. These relations were also robust in both urban and rural areas.

Environmental factors were positively and significantly associated with the probability of impoverishment due to health spending. Households with inadequate housing (UBN 1), overcrowding (UBN 2), lack of sanitation services, or the use of traditional fuels for cooking (coal, firewood, or dung) were associated with increases of 0.22, 0.15, 0.09, and 0.55 percentage points, respectively, in the probability of impoverishment. As in the previous dimensions, these relations were significant in both urban and rural households, suggesting that structural deprivations are related to greater exposure to health risks and, consequently, higher out-of-pocket expenses that may be linked to poverty.

4 Discussion

The results show that impoverishment due to out-of-pocket health expenditures has increased progressively over the last twelve years, from 2.0% of non-poor households in 2010 to 3.6% in 2021, with a peak of 4.6% during the COVID-19 pandemic. This sustained increase coincides with international evidence documenting the impoverishing effect of direct payments for health care, especially in contexts where public financing is insufficient and households must absorb a considerable portion of the cost of care (Akazili et al., 2017; Hooda, 2017; Van Doorslaer, 2007). Using the methodology proposed by Xu (2005), it is confirmed that medical expenses are a central mechanism of downward mobility: even households that were not initially poor crossed the poverty line as a direct result of the need for medical care.

The magnitude of the phenomenon is consistent with what has been observed in various low- and middle-income countries. In India, for example, out-of-pocket spending pushes more than 50 million people below the poverty line each year, mainly due to the cost of medicines and hospitalizations (Garg and Karan, 2009; Hooda, 2017; Keane and Thakur, 2018). In Ghana, out-of-pocket payments increased the poverty rate by between 3.8% and 9.4%, depending on the poverty line used, while the poverty gap increased by up to 42% after accounting for health expenditure (Akazili et al., 2017). Similar results have been documented in Sudan, where the number of poor households grew by between 8% and 17% after health expenditure, highlighting that medical costs displace resources allocated to education, food, and other essential goods (Ebaidalla and Ali, 2019). Even in countries with free or subsidized services, such as Sri Lanka and Uganda, high levels of catastrophic expenditure persist, associated with indirect costs or the expansion of private care (Sisira Kumara and Samaratunge, 2019; Yap et al., 2018).

The pattern found in Peru—higher levels of impoverishment among rural households, those headed by men, and those located in the highlands and jungle—coincides with that observed in India, Ghana, and Indonesia, where limited health infrastructure, greater distance to facilities, and socioeconomic gaps increase financial vulnerability to medical expenses (Barasa et al., 2017; Hafidz et al., 2023; Lavilla, 2012). In these contexts, rural households not only face higher disease prevalence, but also transportation costs, lower availability of generic drugs, and less effective insurance rates.

Likewise, the finding that members of the Comprehensive Health Insurance (SIS) program experience higher levels of impoverishment than those insured by EsSalud reveals significant differences in effective coverage and financial protection between the two systems. This result is consistent with international and regional evidence showing that schemes targeting vulnerable populations, despite expanding access, tend to offer limited protection against medical and pharmaceutical costs, while contributory schemes are better able to reduce the financial risk associated with healthcare (Ku et al., 2019; Li et al., 2012; Petrera Pavone and Jiménez Sánchez, 2018). In countries such as China and Taiwan, for example, basic public insurance has partially reduced out-of-pocket spending, but its effect is weaker among lower-income households, who continue to face high levels of spending on medicines, transportation, and complementary services.

In relation to the associated factors, socioeconomic variables such as higher levels of education and greater per capita income were significantly related to a lower probability of impoverishment. This result is consistent with international evidence, where human capital and economic capacity act as protective barriers against health shocks by enabling more efficient use of services, better self-care practices, and greater capacity to absorb unexpected expenses (Engelgau et al., 2012; Kien et al., 2016). Likewise, attendance at health centers is associated with a lower risk of impoverishment, which coincides with findings indicating that timely care slows the progression of disease and thus avoids more costly late treatments that often trigger catastrophic spending (Aji et al., 2013; Correa-Burrows, 2012). Similarly, the protection observed among EsSalud insured persons reinforces the evidence that contributory schemes—characterized by greater drug coverage, continuity of treatment, and access to more complex services—effectively reduce exposure to impoverishing expenditure (Woldemichael et al., 2016; Yap et al., 2018).

In contrast, factors associated with a higher probability of impoverishment—such as the dependency ratio, the presence of chronic diseases, disability, and a greater number of recipients—reveal patterns similar to those documented in India, China, Iran, and Sudan, where household epidemiological conditions are consistently reported as key correlates of catastrophic and impoverishing expenditure (Ebaidalla and Ali, 2019; Hooda, 2017; Jiang et al., 2012; Raeissi et al., 2018). The presence of chronic diseases, in particular, emerges as a critical associated factor, as it requires prolonged treatment, recurrent visits, and increased consumption of medicines, elements that tend to transfer costs directly to the household when health systems do not guarantee effective financial coverage. This pattern suggests that the system, rather than cushioning health contingencies, may shift financial risk to households, reinforcing pre-existing conditions of vulnerability. Likewise, the greater probability observed in male-headed households and those with young children is consistent with studies showing that life-cycle stages with higher care burdens are related to greater healthcare needs and, consequently, greater exposure to unexpected expenses (Das et al., 2020; Engelgau et al., 2012).

On the other hand, the positive association between environmental factors—overcrowding, inadequate housing, lack of sanitation, and the use of solid fuels—and impoverishment due to out-of-pocket health expenditure is consistent with evidence from Vietnam, China, and Kenya, where structural environmental conditions are associated with a higher incidence of respiratory, gastrointestinal, and chronic diseases, thereby increasing the likelihood of facing high medical expenses (Jiang et al., 2012; Kien et al., 2016; Kimani et al., 2016). The literature also highlights that adverse environmental exposure may generate cumulative effects on health costs, as documented in Iran with regard to air pollution and household medical expenses (Raeissi et al., 2018). In the case of Peru, the higher prevalence of these conditions in rural and peri-urban areas may partly explain the greater incidence of impoverishment observed in these territories, reinforcing the idea that structural health vulnerability is closely related to the financial risk associated with disease.

The study's findings confirm that out-of-pocket spending continues to be a central mechanism associated with impoverishment in Peru, especially among rural households, those with chronic health conditions, and those with lower levels of human capital. This evidence reinforces the need to strengthen the financial protection of the health system through public policies aimed at consolidating truly universal insurance, with coverage that goes beyond basic care and guarantees comprehensive, timely, and quality services, particularly in MINSA and EsSalud facilities. It is also a priority to promote the use of formal health services, given that the preference for drugstores and pharmacies—motivated by lower accessibility or shorter waiting times—is associated with higher out-of-pocket spending and may expose households to risks arising from information asymmetry and self-medication, with potentially serious consequences. Likewise, the associated factors identified suggest that poverty reduction should also be addressed through improvements in basic household services, such as access to drinking water, sanitation, and reduced overcrowding, in addition to expanding effective coverage for pregnant women and people with chronic diseases.

The main strength of this study lies in the application of the methodology proposed by the World Health Organization to estimate impoverishment derived from out-of-pocket health spending, which allows the use of a widely recognized analytical framework and facilitates the international comparability of the results. Nevertheless, the study also presents limitations that should be considered when interpreting the findings. First, due to the empirical design and the observational nature of the data, it was not possible to establish causal relationships between impoverishment and the factors analyzed; therefore, the results should be understood as statistical associations. Second, the estimates rely on information reported by households in the surveys, which may be subject to measurement errors, underreporting, or recall bias, particularly in variables related to income and health expenditures. In addition, the monetary measurement of impoverishment does not fully capture non-monetary dimensions of well-being or coping strategies used by households, such as indebtedness, asset sales, or postponement of treatment.

Regarding future lines of research, it would be pertinent to deepen the analysis of impoverishment according to the type of care received, the nature of expenditures, and preferences in the use of health services. It is also relevant to explore heterogeneity across population subgroups, such as income levels, geographic areas, demographic characteristics, and family structure. Finally, the use of methodological strategies that allow the identification of causal relationships between associated factors and impoverishment is recommended in order to generate more robust evidence for the design of targeted interventions and more effective public policies.

5 Conclusions

The study shows that impoverishment due to out-of-pocket health expenditures in Peru followed an upward trend over the last twelve years, reaching its highest level during the COVID-19 pandemic period. The proportion of non-poor households that fell into poverty due to medical costs increased from 2.0% in 2010 to 4.6% in 2021, confirming that direct healthcare payments continue to be closely associated with downward economic mobility.

The study also reveals significant territorial and socioeconomic inequalities. Rural households, those headed by men, and those located in the highlands and jungle exhibited higher levels of impoverishment, reflecting patterns consistent with barriers to access, lower availability of services, and unfavorable structural conditions. A comparison between insurance schemes shows that SIS members experienced higher levels of impoverishment than EsSalud members, highlighting gaps in effective coverage and suggesting that enrollment alone is not sufficient to ensure full financial protection.

The factors associated with impoverishment identified in the empirical analysis reinforce this interpretation. Years of education, per capita income, and attendance at health facilities were associated with a lower probability of impoverishment, while the presence of chronic diseases, disability, dependency ratio, and housing deficiencies were associated with a higher probability. These findings indicate that poverty linked to health expenditure is not solely related to limitations in financial coverage, but also to socioeconomic and environmental conditions associated with greater household vulnerability. Consequently, health insurance coverage, although essential, appears insufficient when not accompanied by broader improvements in the social conditions related to health.

The results underscore the importance of comprehensive public policies that combine the strengthening of universal health coverage with specific interventions aimed at reducing structural inequalities. This includes improving the effective coverage of SIS, ensuring the availability of medicines and services without hidden costs, and strengthening the response capacity of MINSA and EsSalud. At the same time, it is essential to address socioeconomic and environmental gaps through investments in drinking water, basic sanitation, adequate housing, and the reduction of overcrowding, factors associated with disease incidence and, therefore, with impoverishing expenditure.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. ChatGPT 5 was used to help organize the writing.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Florent Nkouaga, National Association of Insurance Commissioners, United States

Reviewed by: Tongpu Zhao, Peking University, China

Monserrat Serio, Universidad Nacional de Cuyo, Argentina