How Assured Is Health in Peru? A Cross‐Sectional Analysis of the 2019 National Household Survey

ABSTRACT

Background and Aims

In Peru, disparities in medical services have been reported, but costs have not been assessed based on the different health insurance (HI) available. In this cross‐sectional study, we aimed to determine the differences in the cost of consultations and medicines in Peruvians with different types of HI according to the 2019 National Household Survey (ENAHO), which is a representative dataset of the Peruvian population.

Methods

A multivariate‐regression model was used to analyze the relationship between the expenditure on consultations, medicines, dental services, and other expenses and the type of HI. In this case, the type depends on the provider: (1) Social HI called “Seguro Integral de Salud” (SIS) of the Ministry of Health (MINSA); (2) ESSALUD Social Insurance; (3) Armed Forces; and (4) Private sector. Our regressions include demographic controls, and we declared the corresponding survey design to obtain correct standard errors.

Results

A total of 15,592,649 (47.7%) and 8,159,941 (25%) people had SIS and ESSALUD, respectively. In the first quartile, 77.6% had access to HI through the SIS and 16.1% were uninsured; in the fourth quartile, only 17.3% had access to the SIS. 80% and 6% of people in rural areas have access to SIS and ESSALUD, respectively. In urban areas, there is a higher percentage of people who do not have HI (24% vs. 14% in rural areas) (p < 0.05). We found that a SIS affiliate spends an average of USD 22 (80 Soles) on consultations and USD 85 (306 Soles) on medicine. The ESSALUD affiliate spends USD 36 (130 Soles) on consultations and USD 116 (420 Soles) on medicine.

Conclusions

Our results showed differences in the expenditures on consultations, availability of medicines, and costs among people presenting different types of HI affiliations according to ENAHO 2019, where the SIS covers a wide range of insured persons generating high disbursements.

1. Introduction

The health system encompasses all organizations, institutions, resources, and people whose primary purpose is to improve people's health [1]. The Peruvian health system is fragmented and segmented because it is composed of four subsystems or types: (1) Social Health Insurance (HI) called “Seguro Integral de Salud” (SIS) of the Ministry of Health (MINSA); (2) ESSALUD Social Insurance; (3) Armed Forces and National Police of Peru; (4) Private sector; and each of these subsystems has its own autonomous financing and affiliation mechanisms according to specific population groups based on labor conditions, economic income, situation of poverty or vulnerability, and so forth [2].

Universal access to health services is the basis for an equitable health system. This means that all people should have access to adequate, timely, and quality health services, as well as access to quality, safe, and effective medicines. Finally, it is guaranteed that the use of these services should not expose users to financial difficulties [3]. In fact, a good health system should provide financial protection, not only access to needed health. This is particularly important because about one‐fifth of the Peruvian population is living below the national poverty line. Besides, according to the World Health Organization (WHO), Peru has a Universal Health Coverage (UHC) service coverage index of 74.96 in 2019, below the observed levels in other countries of the region, as Brazil (81.13) or Chile (82.28). This index combines 14 indicators of service coverage into a single summary measure.

In 2009, Law 29344, the Framework Law on Universal HI, established the regulatory framework for access and the functions of regulation, financing, provision, and supervision of insurance [4]. It also establishes the Essential HI Plan (PEAS), which is the minimum list of insurable conditions and procedures for the entire population covered by HI [5]. Regarding financing, MINSA's Comprehensive Health System (SIS) is financed through the general budget, which comes from general taxes. ESSALUD is financed by employee contributions (9% of monthly salary). The private sector is supported by insurance and out‐of‐pocket payments [6].

There are many barriers that prevent health professionals from providing high‐quality medical care. Both research and evaluation are an essential part of achieving universal quality health coverage by identifying opportunities and challenges and assessing progress. More policy and implementation research are needed to better understand these barriers and to drive innovation to develop new approaches to transform the health systems [7].

Therefore, we aimed to determine the differences in spending on consultations, medications, and dental services, among other services, in people with different types of insurance according to the 2019 National Household Survey of Peru (“Encuesta Nacional de Hogares”, ENAHO).

2. Materials and Methods

2.1. Study Design, Database, and Sampling

We conducted a secondary analysis of data obtained from the 2019 National Household Survey (ENAHO) of the National Institute of Statistics and Informatics (INEI). This survey allows us to calculate indicators related to the evolution of poverty, welfare, and living conditions of households in Peru. The sampling type used in the survey is probabilistic, area‐based, stratified, multistage, and independent in each region. In 2019, the number of interviewed households with complete or incomplete information amounts to 34,565. From this amount, 21,493 households are in urban areas and 13,072 households are in rural areas. This survey is representative of the Peruvian population.

2.2. Variables

Variables include type of HI, gender, age range, and area of residence (urban and rural). Also, data from module 400 of the ENAHO were considered, which contains information on people's health, such as birth data, people with a chronic illness or disease, access to HI, and spending on medicines (price in US Dollars or Peruvian Soles), among others. Socioeconomic status was measured as real monthly per capita expenditure, divided into quartiles, where the first quartile is mainly composed of people in poverty and extreme poverty.

Regarding the main outcome variable, defined as the expenditure level on consultations, medicines, dental services, and others, we used detailed information provided by the ENAHO. For most items collected by the survey, the recall period is 4 weeks.

2.3. Statistical Analysis

To estimate the differences in the levels of spending among the insured, the following regression will be estimated:

$$y_i={\alpha }_0+{\alpha }_1{x}_i+\sum _{j=2}^C{\alpha }_j{z}_j+{\epsilon }_i,$$

where $y_i$ is the natural logarithm of expenditure on a specific service, $x_i$ indicates which insurance a person is affiliated with (the base category refers to whether the person is affiliated with the SIS), and ${\epsilon }_i$is the error term. It is worth mentioning that we include controls (z) such as gender, age, and geographic scope. In this notation, C is the number of controls. Therefore, it will be possible to know whether having another type of insurance means more expenditure or less than having SIS. It is worth mentioning that if an individual's insurance is from the Armed Forces/Police, University, or Private School, then it will be considered part of the “Other” category. Likewise, only individuals who are household members and of legal age will be considered. We use the natural logarithm of annualized health expenditure on each item plus one, as [8], to deal with zero values. In that study, the authors used ENAHO 2011 to estimate the effect of social HI targeted to the poor in Peru.

2.3.1. Ethical Aspects

We respected the Helsinki Declaration guidelines [9], but IRB approval was respected throughout this study since the ENAHO database is public.

3. Results

According to data from the 2019 ENAHO, in Peru 15,592,649 (47.7%) people have SIS, 8,159,941 (25%) people have the ESSALUDESSALUD insurance, 811,103 people have private insurance (2.5%), and 806,683 people (2.5%) have another type of HI. Thus, 77.7% of the population in Peru has some type of insurance. Figure 1a shows the distribution of the population by type of insurance.

Figure 1.

Characteristics of health insurance in the Peruvian population according to ENAHO 2019. (A) Frequency according to the type of health insurance in Peru. (B) Type of health insurance by age range.

When differentiated by socioeconomic status, in the first quartile, 77.6% have access to HI through the SIS, and 16.1% have no insurance. In contrast, in the fourth quartile, only 17.3% have access to SIS. We also prove that 42.9% have ESSALUDESSALUD insurance, 8.6% have private insurance, and 24.3% of people have no insurance (Table 1).

Table 1.

Type of health insurance by quartile of expenditure. Data in %.

Access to insurance	Quartile 1	Quartile 2	Quartile 3	Quartile 4
SIS	77.6	57.5	38.6	17.3
ESSALUD	6.0	19.5	31.5	42.9
Private	0.1	0.3	1.0	8.6
Other	0.2	0.7	2.1	6.8
No insurance	16.1	22.1	26.8	24.3
Total	100	100	100	100

HI, according to gender, showed that 51% of women have SIS, as opposed to 45% of men. Both 25% of men and women have ESSALUD. In addition, we found that 25% of men do not have HI, while only 20% of women are not insured. We found that 60% of people under 18 years old have SIS, as opposed to 42% in the case of adults. Also, 26% of adults do not have insurance (12 percentage points more than in the case of minors) (Figure 1b).

In terms of geographic scope, significant differences are found between people living in urban and rural areas. 80% of people in rural areas have SIS, while only 39% in urban areas have it. Regarding ESSALUD, only 6% in rural areas have this type of insurance, as opposed to the 30% of coverage observed in urban areas. In fact, in urban areas, there is 3% coverage in relation to private insurance (in rural areas, coverage is less than 1%). Finally, in urban areas, there is a higher percentage of people who do not have HI (25% against 13% in rural areas). In all cases, we ran bivariate regressions, where dependent variables were different versions of access to HI (access to SIS, none, etc.), and the independent variable was the corresponding demographic variable. We found all coefficients significant at 1%. We considered this, and we ran the main regression below.

On the other hand, it was found that an SIS member spends an average of USD 22.3 (80 Soles) on consultations, USD 85.3 (306 Soles) on medicine, USD 95 (343 Soles) on dental services, and USD 29 (103 Soles) on other expenses. On the other hand, an ESSALUD affiliate spends an average of USD 36 (130 Soles) on consultations, USD 117 (420 Soles) on medicine, USD 237 (854 Soles) on dental services, and USD 61 (221 Soles) on other expenses. The average expenditure on different items by type of HI is shown in Table 2.

Table 2.

Average annual health expenditure by type of insurance. Data in USD (Soles).

	In Peruvian Soles				In US Dollars
	Mean	SD	Min	Max	Mean	SD	Min	Max
Consultations
SIS	80.17	202.40	0	7573	22.27	56.22	0	2104
ESSALUD	130.26	284.88	0	5926	36.18	79.13	0	1646
Privado	321.00	404.32	0	3611	89.17	112.31	0	1003
Otro	196.02	318.82	0	2808	54.45	88.56	0	780
No tiene	204.81	317.14	0	4870	56.89	88.10	0	1353
Total	130.03	270.74	0	7573	36.12	75.21	0	2104
Medicine
SIS	305.95	972.97	0	42,472	84.99	270.27	0	11,798
ESSALUD	420.22	1124.24	0	36,328	116.73	312.29	0	10,091
Privado	693.37	1544.73	0	16,697	192.60	429.09	0	4638
Otro	625.51	1238.30	0	12,965	173.75	343.97	0	3601
No tiene	458.93	1011.54	0	34,087	127.48	280.98	0	9469
Total	394.57	1059.45	0	42,472	109.60	294.29	0	11,798
Dental services
SIS	342.83	905.46	0	14,109	95.23	251.52	0	3919
ESSALUD	854.11	2099.62	0	67,600	237.25	583.23	0	18,778
Privado	1310.61	2259.08	0	16,266	364.06	627.52	0	4518
Otro	1067.47	2236.01	0	20,099	296.52	621.11	0	5583
No tiene	810.64	1644.05	0	17,370	225.18	456.68	0	4825
Total	703.79	1711.81	0	67,600	195.50	475.50	0	18,778
Other expenses
SIS	103.14	255.96	0	7991	28.65	71.10	0	2220
ESSALUD	220.79	547.14	0	48,021	61.33	151.98	0	13,339
Privado	379.73	756.74	0	8024	105.48	210.20	0	2229
Otro	313.33	795.56	0	12,106	87.04	220.99	0	3363
No tiene	179.57	381.05	0	12,200	49.88	105.85	0	3389
Total	171.19	441.37	0	48,021	47.55	122.60	0	13,339

Source: INEI‐2019 National Household Survey (ENAHO). *Only adults are considered.

Table 3 shows the relationship between annual expenditure on consultations, medicines, dental services, and other expenses and the type of insurance. To do so, we used a linear regression model of health expenditure (in log) as a function of HI status and other covariates. Specifically, we controlled for a series of individual‐level characteristics such as gender, geographic area, domain, and age. In particular, there is a significant heterogeneity between type of expense and type of insurance. For example, ESSALUD insured spend 9% more than SIS in terms of spending on consultations. This coefficient rises to 31% in the case of dental services and 29% in other expenses. Likewise, those who do not have insurance spend 43% and 40% more on consultations and medicines than those insured with the SIS, respectively. On the other hand, when analyzing the control variables, it is found that in all cases women and those who live in urban areas spend more than men and those who reside in rural areas, respectively.

Table 3.

Relationship between health spending and type of insurance.

	(1)	(2)	(3)	(4)	(5)
	Consultations	Medicines	Dental services	Other expenses	All
ESSALUD	0.0867**	−0.0572	0.311***	0.288***	0.357***
	(0.0276)	(0.0381)	(0.0237)	(0.0330)	(0.0433)
Private	0.413***	0.223*	0.836***	0.740***	1.146***
	(0.0874)	(0.110)	(0.0886)	(0.107)	(0.128)
Other	0.173**	0.371***	0.529***	0.462***	0.911***
	(0.0639)	(0.0946)	(0.0654)	(0.0799)	(0.101)
No insurance	0.432***	0.401***	0.220***	0.0259	0.478***
	(0.0263)	(0.0350)	(0.0205)	(0.0274)	(0.0386)
Female = 1	0.114***	0.261***	0.0516**	0.425***	0.478***
	(0.0193)	(0.0241)	(0.0166)	(0.0204)	(0.0266)
Age	0.00113*	0.0188***	−0.00410***	0.0124***	0.0202***
	(0.000550)	(0.000766)	(0.000433)	(0.000612)	(0.000806)
Urban = 1	0.301***	0.575***	0.173***	0.104***	0.640***
	(0.0247)	(0.0366)	(0.0167)	(0.0308)	(0.0432)
Central Coast	0.0378	0.566***	0.0517	−0.201***	0.360***
	(0.0495)	(0.0647)	(0.0320)	(0.0478)	(0.0718)
South Coast	−0.480***	−0.573***	0.0869*	−0.360***	−0.599***
	(0.0459)	(0.0656)	(0.0387)	(0.0494)	(0.0770)
North Highlands	−0.0241	−0.157*	0.141***	0.203**	0.0867
	(0.0511)	(0.0727)	(0.0347)	(0.0621)	(0.0843)
Central Highlands	−0.0933*	−0.190***	0.0996***	0.183***	−0.00909
	(0.0417)	(0.0554)	(0.0287)	(0.0471)	(0.0643)
South Highlands	−0.356***	−0.639***	0.156***	−0.0997*	−0.498***
	(0.0422)	(0.0579)	(0.0314)	(0.0471)	(0.0711)
Amazon rainforest	−0.257***	−0.177***	0.0689**	0.106*	−0.0600
	(0.0386)	(0.0481)	(0.0255)	(0.0440)	(0.0590)
Lima	−0.147***	0.0682	0.126***	0.297***	0.229***
	(0.0435)	(0.0579)	(0.0300)	(0.0515)	(0.0687)
Constant term	0.537***	0.418***	0.212***	0.177***	1.009***
	(0.0466)	(0.0614)	(0.0325)	(0.0510)	(0.0703)
N	82,223	81,815	82,254	82,139	82,265
R 2	0.020	0.046	0.021	0.041	0.055

Source: INEI‐2019 ENAHO. Only adults are considered. Robust standard errors are in parentheses.

^* p <0.05

^** p <0.01

^*** p <0.001.

4. Discussion

This study found that 16% of people in the first quartile do not have insurance compared to 24% of people in the fourth quartile, showing that people in the fourth quartile, since they have greater purchasing power, are more likely to be able to finance the corresponding expenditure on consultations and medicines when they need medical care compared to people in the first quartile, most of whom are poor or extremely poor.

Previously, in a secondary analysis of the 2017 ENAHO, they found that 56.5% of older adults reported out‐of‐pocket spending on health, which generated inequity for health access, especially in economically and socially vulnerable groups [10]. Out‐of‐pocket payment implies a direct threat to a risk of impoverishment, which can be enhanced when the disease suffered is of high cost, such as cancers, rare diseases, and even chronic diseases, and the disease by the new 2019 coronavirus (COVID‐19) [11]. The pandemic has shown the downsides of obtaining employer‐based HI, with out‐of‐pocket payment resulting in increased care costs during the lockdown [12]. Despite COVID‐19's shock to governments around the world, only a handful of countries have proposed pandemic risk insurance schemes to protect insurance businesses [13]. The countries of the region have not only not contemplated these proposals, but also government technological investment has not prioritized this item during the pandemic [14].

Regarding gender, it was found that 25% and 20% of men and women do not have HI, respectively. This apparent difference in coverage when differentiating by gender may be linked to the age ranges considered. For example, in the 0–17 age range, the percentage of men and women with SIS, ESSALUD, and uninsured was 60%, 23%, and 14%, respectively. In contrast, when the sample is restricted to adults (≥ 18 years old), the percentage of women with SIS is 47% (38% for men), 30% of men are uninsured (22% for women), and there are no significant differences in the ESSALUD insurance coverage. The results found are consistent with the research by Mezones‐Holguín et al. (2019) who, in their study on HI coverage using annual data from the 2009–2017 ENAHO, found that by 2017, 49.7% and 26.2% of women were covered by SIS and ESSALUD, respectively [15].

It was found that 60% of people < 18 years old had SIS compared to 42% of adults. This is mainly explained by the progressive increase in SIS coverage and the subsequent search for universalization, which initially focused on vulnerable population groups, such as pregnant women, infants, and adolescents. Likewise, it should be remembered that the SIS arose from the unification of the Maternal and Child Insurance and the Free School Insurance. The data found in this study correlate with previous MINSA figures, as 83.3% of the population under 18 years old had at least one type of insurance by 2016, with SIS having the highest coverage (58.1%), followed by ESSALUD with 22%. These figures have increased over the years, since by 2004 the coverage of population under 18 years old was 37.4% and 13.5% for SIS and ESSALUD, respectively [16].

Our findings showed that dramatically 80% of people in rural areas have SIS, while only 39% in urban areas have it. This result reflects the targeting criteria, since it concentrates on those living in poverty, mostly in rural and peri‐urban areas. However, it should be taken into account that it is precisely in these areas that the supply of health facilities is most precarious. There are several factors that influence the search for care, such as the supply of health facilities and the risk level of the individual, among others [17, 18].

The results shown indicate that the SIS does indeed cover the healthcare needs of its members to some extent, but that such coverage is not comprehensive, since the beneficiaries actually end up financing such care from their own income sources. In fact, of the 3,158,136 SIS insured who received medical consultations, 47% (1,405,862 people) had to pay for this care out of their own income. This means that in practice, the SIS did not end up being comprehensive for almost half of the insured who received medical consultations. This can be explained by the quality of health supply in the country since rural areas tend to have a higher proportion of health posts and a scarce supply of medicines [19].

Finally, from the above result, it can be seen that, in order for the SIS to be comprehensive in relation to the concept of medical consultation, it should allocate a budget of USD 71,193,767 (256,297,561 Soles) for this purpose, which represents 15% of its 2019 Modified Institutional Budget. If the concept of medicines were to be added to this, then it would represent 92% of this budget for 2019. It is important to mention that this exercise only tries to provide a first overview of how much budget would be required to allocate to the SIS to guarantee a minimum quality coverage throughout the country and that it does not represent an additional cost for Peruvian households, mainly those living in poverty.

This study had limitations. Our analysis has used the official ENAHO 2019 report; however, there may be recent updates that may have important changes regarding HI. Another limitation is the set of components not considered in the model (i.e., Human Development Index and poverty quintiles). Demographic, economic, and social factors must be factored in so that inequalities between regions can be seen [20]. Finally, this analysis has been carried out before the COVID‐19 pandemic, and it is possible that there will be dramatic changes in access to health between communities. The pandemic has not only affected the economy [21], it has also changed the course of the disease [22]. It is important to account for this variable in future studies because changes in cost and high demand for health care can affect access to health, creating more disparities between communities.

In conclusion, this study has demonstrated differences in spending on consultations, availability of medicines, auxiliary examinations, and costs among people who have different types of insurance according to the 2019 ENAHO, evidencing a broad coverage of the SIS that covers the immeasurable minority of care costs and therefore reduces its comprehensiveness. It is necessary that these reluctances in Peruvian health programs are positioned in a current context of sustainability and that they take advantage of the current health visibility due to COVID‐19, to reorganize and universalize themselves in an organized manner.

Author Contributions

Hans Contreras‐Pulache: conceptualization, data curation, formal analysis, methodology, project administration, validation, writing – original draft, writing – review and editing. Víctor Quispe: conceptualization, data curation, formal analysis, methodology, software, writing – original draft. Gloria Cruz‐Gonzales: investigation, resources, supervision, writing – review and editing. Jeel G Moya‐Espinoza: methodology, supervision, validation, Writing – original draft. William Cruz‐Gonzales: resources, supervision. Jeel Moya‐Salazar: data curation, investigation, resources, validation, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

1. Transparency Statement

Jeel Moya‐Salazar affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

Supporting information

STROBE‐checklist‐v4‐cross‐sectional.

Data Availability Statement

Jeel Moya‐Salazar confirms that he has full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.