Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Primary Health Care Utilization by the Mexican Indigenous Population: The Role of the Seguro Popular in Socially Inequitable Contexts

Primary Health Care Utilization by the Mexican Indigenous Population: The Role of the Seguro Popular in Socially Inequitable Contexts

  • Rene Leyva-Flores, 
  • Edson Servan-Mori, 
  • Cesar Infante-Xibille, 
  • Blanca Estela Pelcastre-Villafuerte, 
  • Tonatiuh Gonzalez
PLOS
x

Abstract

Objective

To analyze the relationship between primary health care utilization and extended health insurance coverage under the Seguro Popular (SP) among Mexican indigenous people.

Methodology

A cross-sectional analysis was conducted using data from the Mexican National Nutrition Survey 2012 (n = 194,758). Quasi-experimental matching methods and nonlinear regression probit models were used to estimate the influence of SP on primary health care utilization.

Results

25% of the Mexican population reported having no health insurance coverage, while 59% of indigenous versus 35% of non-indigenous reported having SP coverage. Health problems were reported by 13.9% of indigenous vs. 10.5% of non-indigenous; of these, 52.8% and 57.7% respectively, received primary health care (p<0.05). Economic barriers were the most frequent reasons for not using primary health care services. The probability of utilizing primary health care services was 11.5 percentage points higher (p<0.01) for indigenous SP affiliates in comparison with non-indigenous, in similar socioeconomic conditions.

Conclusion

Socioeconomic conditions, not ethnicity per-se, determine whether people utilize primary health care services. Therefore, SP can be conceived as a public policy strategy which acts as a social buffer by enhancing health care utilization regardless of ethnicity. Further analysis is required to explore the potential gaps as a result of SP coverage among socially vulnerable groups.

Introduction

Worldwide, access to health care for indigenous groups has always been limited due to social inequity [1][3]. Although representing 5% of the world's population, indigenous peoples account for approximately 15% of global poverty [4] and one-third of extreme poverty in rural areas [4].

Mexican indigenous people (∼7% of the Mexican population in 2010) [5] are characterized by steep marginalization levels and wide social gaps [6], [7]. In 2010, 44.2% of those residing in indigenous municipalities [8] suffered from extreme poverty (according to the Mexican National Council for the Evaluation of Social Development Policy or CONEVAL, extreme poverty is a condition characterized by such meager incomes that, even when destined entirely to purchasing food, they prove insufficient for acquiring the necessary nutrients to lead a healthy life); 78.6% earned less than three minimum wages; 83.3% were below the well-being threshold or under an income equivalent to the total value of the consumption food and non-food bundle per person per month (according to CONEVAL at April 2013: US$122.1 in localities with less than 2500 inhabitants or rural and US$190.3 and localities with more than 2,500 inhabitants or rural), and 75.6% lived in marginalized communities [6]. Additionally, in 2010, the infant mortality rate was 63% higher among indigenous versus non-indigenous municipalities, and prevalence of stunting was threefold in indigenous in comparison with non-indigenous communities [9]. These figures relate directly to the living conditions of indigenous peoples and the countless obstacles they face in accessing primary health care services.

The Mexican government has responded by stepping up social policies that seek greater access to education, food and health [10], particularly for populations stricken by acute social vulnerability, such as the indigenous communities.

Since 2003, Seguro Popular (SP) has constituted one of several social health protection strategies created to remove economic barriers and improve access to health care [11], with the SP program providing public funds for populations in the first income deciles who lack social security [11][14]. Since its inception, public resources allocated to SP have grown 13.3 times, that is, from $382.6 million USD in 2004 to $5,087.6 million USD in 2012 ($4,315.3 million MXN-2004- vs. $67,004 million MXN-2012, according to BANXICO, average monthly peso/dollar exchange rates were 11.28 in 2004 and 13.17 in 2012), when the annual average reached $100 USD per affiliate [15].

Seguro Popular impact assessment studies have demonstrated its protective effect against catastrophic health spending among its beneficiaries [12][14], [16][18], and analyses on health care utilization have sighted a number of contributions [19], [20]. However, evidence regarding performance in rural indigenous communities at the lowest economic levels [21], [22] suggests that results have been less favorable among these beneficiaries. SP coverage increased from 14% to 61.9% for the indigenous population from 2006 to 2012, but from 10% to 35.7% for the non-indigenous population throughout the ten years following its implementation [23]. In order to establish the relationship between primary health care utilization and extended insurance coverage for Mexico's poorest, specifically for the indigenous population, this study analyzed the role of SP as a socioeconomic buffer against barriers to health care services in 2012.

Methodology

A cross-sectional study was conducted with data from the Mexican National Nutrition Survey (ENSANUT) 2012, a probabilistic survey providing evidence not only at national and state levels, but also by urban and rural strata [24]. Data from the survey's household module were retrieved specifically with regard to socio-demographic indicators, health status and primary health care utilization. This survey was approved by the Research and Ethics Committees of the National Institute of Public Health. ENSANUT was applied to 194,758 persons in 50,528 households, excluding individuals who provided incomplete socio-demographic data (0.74%); already benefited from private health insurance (0.34%), or lacked information on morbidity, type of health problem, or attendance at primary health care services (0.4%). With a sampling loss of 1.5%, analyses were performed on a final sample of 191,849 individuals (N = 113,039,438).

The variable of interest was primary health care utilization. This is an indicator that has proved sensitive to shifts in economic capacities influencing the regulation of health care demand [25][27]. Congruent with previous studies, primary health care utilization was defined according to the respondents' self-reported health problems - both experienced within two weeks prior to ENSANUT, and treated by medical personnel on an outpatient basis [28]. People who reported being attended to by non-medical personnel (13.9%) were classified as non-users.

The indigenous population was identified according to the definition proposed by the National Commission for the Development of Indigenous Peoples (CDI for its initials in Spanish), whereby a household is considered indigenous if the head of the family, a spouse and/or an ascendant self-identifies as a speaker of an indigenous language [29]. 24,090 people (∼10.8 million individuals, or 9.6% of the total Mexican population) were thus identified as indigenous.

Analyses were performed on individual and community-related factors linked by former studies [28], [30][34] to the likelihood of receiving health care and/or utilizing health services. Individual variables included: gender, age, education, employment, marital status, number of household members, annual spending (per resident per capita) and receipt of cash transfers from the Mexican Oportunidades anti-poverty program. Places of residence included locality sizes: (i) rural (<2,500 inhab.); (ii) urban (2,500–100,000 inhab.), and (iii) metropolitan (>100,000 inhab.). Lastly, marginalization levels were classified as: (i) very low/low; (ii) medium, and (iii) high/very high [35].

In terms of processed data, the study began by describing the socio-demographic characteristics of the indigenous/non-indigenous groups. Statistically different group traits were obtained by applying statistical independence tests (χ2 for categorical variables and t-student for continuous variables). Additionally, health assurance categories were determined as: (i) having Social Security (SS), (ii) having SP, and (iii) having no SS or insurance whatsoever.

The SS population was then excluded from the study sample to achieve more homogeneous groups, given that 86% of the indigenous population reported having no SS. At this point, the analytical sample was circumscribed exclusively to SP versus non-SP subjects (63% of the total population). Analyses were then performed on primary health care utilization for the self-reported health problems mentioned above (defined as chronic, acute or others) [33]. In the case of problems unattended to by health personnel, impediments cited by respondents were classified as: (i) factors related to service demand (lack of money); (ii) factors related to service supply (ie. lack of confidence, ill-treatment, unavailability and remoteness), and (iii) other causes. Results were then estimated for the sample population, bearing in mind the impact of design on the survey.

Assessing the impact of being indigenous on the likelihood of obtaining primary health care was subject to confounding factors and potential differences inherent in a self-selected study sample marked by the nonrandom assignment of indigenous subjects and the non-experimental design of ENSANUT. Propensity score matching (PSM) was thus applied, as it is one of the methods most frequently used for recreating experimental conditions and obtaining causal inference. PSM allows constructing an appropriate comparison group vis-à-vis the treatment group, as well as controlling for biases indicated by observable variables [36][39].

The indigenous/non-indigenous populations were thus matched according to the co-variables described above, using the one-to-one nearest neighbor algorithm [39]. This and further analyses on matching confirmed statistical similarity between the two groups and bias correction. Non-linear probability probit models [40] were then applied to the matched sample to assess how the attainment of primary health care from medical personnel (adjusted by PSM) was influenced, not only by ethnicity, but also by the interaction of ethnicity with SP affiliation. Marginal effects (in percentage points or pp) and 95% confidence intervals were obtained. All analyses were carried out with STATA SE v13.1 software [41].

Results

Table 1 illustrates the socio-demographic characteristics of the indigenous/non-indigenous study populations. While men and women were distributed proportionately between the groups, indigenous respondents belonged to a younger age bracket (0–19 yrs., 45.3%) than their non-indigenous counterparts (20–49 yrs., 43.3%), and presented lower education levels, participation in the labor market, and per capita spending (∼$68.1 USD/month vs. $128.9 USD/month). Additionally, significantly more indigenous (44.1%) than non-indigenous (14.1%) respondents occupied Quintile I of per capita spending, belonged to Oportunidades households (58.2% vs. 22%), and dwelled in both rural localities (53.5% vs. 19.4%) and highly marginalized areas (71.3% vs. 17%).

thumbnail
Table 1. Socio-demographic conditions of the indigenous and non-indigenous Mexican population, 2012.

https://doi.org/10.1371/journal.pone.0102781.t001

On comparing indigenous with non-indigenous health insurance enrollment (Table 2), the former proved lower (2.8 times) with SS, but higher with SP (58.7% and 34.5%, respectively). Furthermore, 27% of the indigenous and 25.5% of the non-indigenous study population reported having no health insurance whatsoever.

thumbnail
Table 2. Health Insurance among indigenous and non-indigenous Mexican population, 2012.

https://doi.org/10.1371/journal.pone.0102781.t002

Of the total number of individuals affiliated (or not) to SP who reported health problems in the two weeks prior to ENSANUT, 13.9% were indigenous and 10.5% non-indigenous. Also, slight differences were observed with regard to health problems (Figure 1). Indigenous people suffered more from acute problems while their non-indigenous counterparts suffered more from chronic problems (Panel A). Additionally, fewer indigenous (52.8%) than non-indigenous subjects (57.7%) received any kind of health care (Panel B). Lastly, 10.6% of indigenous against 4.1% of non-indigenous attended SS facilities. 57.4% of indigenous and 42.6% of non-indigenous subjects received health care by the Ministry of Health (SSA according to its initials in Spanish) while 32% of indigenous and 53.3% of non-indigenous people utilized private physicians (Panel C). Regarding those who did not utilize primary health care services, 59.7% of non-indigenous and 55.9% of indigenous respondents cited reasons associated with lack of money as the main barrier. Other reasons referred to service supply (37.6%) and perception of services (40.9%), including lack of confidence, poor treatment, unavailability and remoteness. No significant differences were observed between the groups in this regard (Panel D).

thumbnail
Figure 1. Indigenous population: morbidity, type and place of healthcare, and reasons for not using primary health care services.

A. Type of health problem. B. Type of assistance received for reported health problem. C. Place of assistance for reported health problem. D. Reasons for not receiving assistance for reported health problem. Source: Mexican National Nutrition Survey, 2012. Note: Estimates take into account the effect of survey design. Social security beneficiaries were excluded.

https://doi.org/10.1371/journal.pone.0102781.g001

Considering similar socioeconomic conditions, Table 3 shows the estimated impact of ethnicity on the probability of primary health care utilization. In general, Model 1 yielded a 5.69 pp higher probability for indigenous versus non-indigenous subjects. However, after including SP enrollment, Model 2 yielded the same probability for both populations, but a 10.2 pp higher probability for SP beneficiaries. Lastly, after adding the interaction between the two variables, Model 3 yielded an 11.5 pp higher probability of primary health care utilization for individuals who were both indigenous and SP affiliates. The variables proved statistically insignificant when analyzed independently.

thumbnail
Table 3. Impact of being indigenous on the probability of using primary health care services.

https://doi.org/10.1371/journal.pone.0102781.t003

Discussion

Social inequity has been a persistent problem among ethnic groups throughout the world [1][3]. Mexico, a country with an array of indigenous communities historically marked by deep social gaps, is not spared from this reality. To overcome social inequity and improve the living conditions of vulnerable groups, namely the indigenous communities [11], [42], the Mexican government has implemented a host of initiatives. For example, in the early 1940's there was an emergence of social development programs, some of which followed an “assistentialist” approach and have remained untouched in their content since then (see catalog of Federal Social Development Programs and Actions published by the CONEVAL). In 2011, only 14 (5.1%) out of 273 Federal Programs and Actions for Social Development were aimed specifically at the indigenous population, representing a scant 1.4% of total financial resources allocated to national social policy, or ∼0.1% of GDP (Federal Programs and Actions for Social Development were budgeted $730,581.47 million MXN). Moreover, although 40% of the 273 programs related to health care, none were targeted to the indigenous population (see Social Policies in Mexico: Progress and Challenges in 2012, CONEVAL).

Currently, SP is one of the most ambitious social programs undertaken by the Mexican government. It aims to provide financial protection regarding health for the poorest and non-Social Security insured. It also aims to provide opportune and quality access to medical, surgical, pharmaceutical and hospital services required to comprehensively meet the health needs of its beneficiaries [43]. This study focused on primary health care utilization by indigenous people, underscoring the role of SP in facilitating these services in acutely inequitable contexts. Based on quasi-experimental methods, our findings endorse the hypothesis that SP offsets the barriers preventing the use of primary health care services by indigenous and non-indigenous in similar socioeconomic conditions. Our findings suggest that it is not being indigenous per se, but rather the lack of financial assurance for accessing health care, that hinders primary health care utilization. In principle, from this perspective, SP can be conceived as a public policy strategy that acts as a buffer by enhancing primary health care utilization regardless of ethnicity. However, given the persistence of countless uninsured Mexicans [22], the SP social policy may be having an adverse effect by widening the gap between those who utilize and not health services in the poorest population segments. In other words, having or not SP benefits may actually be shaping a new social gap within the vulnerable population segments. Socio-political outlooks on this matter span the gradualist spectrum. Some outlooks have the hope that universal coverage will eventually occur and contribute to social equity in health, while public statements from recent governments [11] affirm that this goal has already been met. To the contrary, however, evidence published by Laurell [22], the National Council for the Evaluation of Social Development Policy (CONEVAL for its initials in Spanish) [12] and, more recently, by ENSANUT 2012, demonstrates that as many as 21.4% of the Mexican population do not benefit from any health insurance whatsoever [44].

While recognizing that “SP has achieved 51.8 million affiliates,” CONEVAL reports that “there is a population group that turns elsewhere for its health-related financial risk management” [12]. For instance, in the case of indigenous communities, one out of ten individuals is covered by the Mexican SS system, a mechanism for providing expanded services, that is, services unrestricted to a catalog such as the one established under SP financing.

What exactly is the contribution of the SP financial insurance plans to socially vulnerable groups? Studies have identified that their greatest contribution lies in the reduction of household catastrophic health spending [12], [13], although exclusively for the health problems listed under the General Health Services Catalog (CAUSES for its initials in Spanish) [22]. CONEVAL [12] and Laurell [22] have indicated the following shortcomings in SP's organization as being potentially responsible for the observed gaps in primary health care utilization: inadequate distribution of physicians, insufficient accreditation mechanisms to ensure quality care, long wait times, and limited access to information on the rights of beneficiaries. According to CONEVAL, the claims in the official SP publications differ from the public perception, particularly as regards effective access and service quality” [12].

Finally, this study did not intend (and was therefore not designed) to assess either the effect of SP on population health status or the impact of SP organization on health care utilization. Studies with a distinct focus on these variables are required to meet the growing interest in the results of government programs for social development and poverty reduction. The ensuing analyses and existing evidence on these programs would contribute to a deeper understanding of their scope, functionality and role in public well-being.

Author Contributions

Conceived and designed the experiments: RL ESM CIX BEPV TG. Analyzed the data: ESM. Wrote the paper: RL ESM CIX BEPV TG.

References

  1. 1. Hall G, Patrinos H (2006) Indigenous Peoples, Poverty and Human Development in Latin America: 1994-2004. Hall, Gillette Patrinos, Harry Anthony edn. Great Britain: Palgrave Macmillan Publishers; 2005.
  2. 2. Partridge W, Uquillas J, Johns K (1998) Including the Excluded: Ethnodevelopment in Latin America Description. In Poverty and Inequality: Proceedings of the Annual World Bank Conference on Development in Latin America and the Caribbean. Washington, DC: The World Bank 1998: 229–250.
  3. 3. Plant R (1998) Issues in Indigenous Poverty and Development, IDB Working Paper [http://www.iadb.org/sds/doc/952eng.pdf.].
  4. 4. International Fund for Agricultural Development (IFAD) (2007) Statistics and key facts about indigenous peoples. Rome: IFAD. Available: http://www.ruralpovertyportal.org/web/guest/topic/statistics/tags/indigenous%20peoples
  5. 5. Instituto Nacional de Estadística Geografía e Informática (2010) Principales resultados del Censo de Población y Vivienda del 2010. Available: http://www.inegi.org.mx/default.aspx. Accessed 2013 June 17.
  6. 6. Handa A (2007) Indigenous Health in Mexico. In Racial and Ethnic Disparities in Health in Latin America and the Caribbean. Edited by Giuffrida A, Bernal R, Cárdenas M, Handa A, Trujillo AJ, Vernon JA, Rodriguez Wong L, Angeles G, Mayer-Foulkes D, Larrea C. Washington, D.C.: Inter-American Development Bank 2007: 65–93.
  7. 7. Consejo Nacional de Población (CONAPO) (2013) República Mexicana: Indicadores demográficos de la población indígena, 2000–2010. Available: http://www.portal.conapo.gob.mx/index.php?option=com_content&view=article&id=37&Itemid=235. Accessed 2013 June 17.
  8. 8. Comisión Nacional para el Desarrollo de los Pueblos Indígenas (CDI) (2006) Regiones indígenas de México. Comisión Nacional para el Desarrollo de los Pueblos Indígenas, Programa de las Naciones Unidas para el Desarrollo; coord. Enrique Serrano Carreto. México: CDI: PNUD, 2006.
  9. 9. Berrío L, Reyes N (2008) Las mujeres guerrerenses y la muerte materna. En: Coalición por la Salud de las Mujeres. La Salud de las Mujeres Guerrerenses. Retos Legislativos. México, 2008.
  10. 10. The World Bank (WB) (2012) México: Panorama de la Protección Social. Washington, D.C.: The World Bank; 2012.
  11. 11. Knaul F, González-Pier E, Gómez-Dantés O, García-Junco D, Arreola-Ornelas H, et al. (2012) The quest for universal health coverage: Achieving social protection for all in Mexico. The Lancet, 380(9849): 1259–1279.
  12. 12. Consejo Nacional de Evaluación de la Política de Desarrollo Social (CONEVAL) (2013) Informe de evaluación específica de desempeño 2012-2013: Seguro Popular. Available: http://www.coneval.gob.mx/Evaluacion/Eval_Especificas/Paginas/EED_2012-2013/EED_2012_2013_SALUD.aspx.
  13. 13. King G, Gakidou E, Imai K, Lakin J, Moore R, et al. (2009) Public policy for the poor? A randomised assessment of the Mexican universal health insurance programme. The Lancet, 373(9673): 1447–1454.
  14. 14. Knaul F, Arreola-Ornelas H, Méndez-Carniado O, Bryson-Cahn C, Barofsky J, et al. (2006). Evidence is good for your health system: Policy reform to remedy catastrophic and impoverishing health spending in Mexico. The Lancet, 368(9549): 1828–1841.
  15. 15. Sistema de Protección Social en Salud (2013) Informe de resultados del SPSS Enero-Diciembre 2012. Secretaría de Salud. Available: http://www.seguro-popular.salud.gob.mx/index.php?option=com_content&view=article&id=277&Itemid=293. Accessed 2013 June 17.
  16. 16. Galárraga O, Sosa-Rubí S, Salinas-Rodríguez A, Sesma-Vázquez S (2010) Health insurance for the poor: Impact on catastrophic and out-of-pocket health expenditures in Mexico. The European Journal of Health Economics, 11(5): 437–447. Available: http://www.jstor.org/stable/40963266.
  17. 17. Sosa-Rubí S, Salinas-Rodríguez A, Galárraga O (2011) Impacto del seguro popular en el gasto catastrófico y de bolsillo en el México rural y urbano, 2005–2008. Salud Pública De México, 53: 425–435.
  18. 18. Wirtz V, Santa-Ana-Tellez Y, Servan-Mori E, Avila-Burgos L (2012) Heterogeneous effects of health insurance on out-of-pocket expenditure on medicines in Mexico. Value in Health, 15(5), 593–603.
  19. 19. Sosa-Rubí S, Galárraga O, Harris J (2009) Heterogeneous impact of the “Seguro popular” program on the utilization of obstetrical services in Mexico, 2001–2006: A multinomial probit model with a discrete endogenous variable. Journal of Health Economics, 28(1): 20–34.
  20. 20. González-Block M, Sauceda-Valenzuela A, Santa-Ana-Téllez Y (2010) Factores asociados a la demanda de servicios para la atención del parto en México. Salud Pública de México, 52(5): 416–423.
  21. 21. Grogger J, Arnold T, León A, Ome A, Triyana M. (2012) Identificación y análisis de los efectos del Seguro Popular en el gasto en salud de los afiliados. CIDE. Unpublished manuscript.
  22. 22. Laurell AC (2013) Impacto del seguro popular en el sistema de salud mexicano. Buenos Aires, Argentina: CLACSO, 2013.
  23. 23. Leyva-Flores R, Infante-Xibille C, Gutiérrez-Reyes JP, Quintino-Pérez F (2013) Inequidad persistente en salud en los pueblos indígenas de México, 2006-2012: retos para el sistema de protección social. Salud Publica Mex 2013 55: S1–23-S128.
  24. 24. Gutierrez JP, Rivera-Dommarco J, Shamah-Levy T, Villalpando-Hernández S, Franco A, et al.. (2012) Encuesta Nacional de Salud y Nutrición 2012. Resultados Nacionales. Cuernavaca, México: Instituto Nacional de Salud Pública (MX), 2012.
  25. 25. Martinez G, Aguilera N, Chernichovsky D (2009) The Mexican health care system and the emerging paradigm in modern systems. Department of Health Systems Management. Faculty of Health Science. Ben-Gurion, University of the Negev, 2009.
  26. 26. Molina-Rodríguez J, Poblano-Verastegui O, Carrillo-Díaz C, Saucedo-Valenzuela (2006) A Utilización de servicios de salud en México. Salud Tabasco; 2006, 12: 427–432.
  27. 27. Céspedes-Londoño J, Jaramillo-Pérez I (2002) Impacto de la reforma del sistema del sistema de seguridad social sobre la equidad en los servicios de salud en Colombia. Cad. Saude Publica; 2002, 18 (4): 1003–1024.
  28. 28. Danese-De los Santos L, Valencia-Mendoza A, Sosa-Rubí S (2011) Analysis of Changes in Health-Care Service Access and Choice of Provider in Mexico: The Utilization of Curative Health Services 2000–2006. BMC Public Health 2011 11: 771
  29. 29. Comisión Nacional para el Desarrollo de los Pueblos Indígenas (CDI) (2012) Los hogares y la población indígena. Available: http://www.cdi.gob.mx/index.php?id=211&option=com_content&task=view, 2009. Accessed 2012 October 1.
  30. 30. Brown C, Pagán J, Rodríguez-Oreggia E (2005) The decision-making process of health-care utilization in Mexico. Health Policy 2005 72: 81–91.
  31. 31. López-Ceballos D, Chunhuei C (2009) Health-care utilization in Ecuador: a multilevel analysis of socio-economic determinants and inequality issues. Health Policy and Planning 2009: 1–10.
  32. 32. Valdivia M (2002) Public health infrastructure and equity in the utilization of outpatient health-care services in Peru. Health Policy and Planning 2002 17: 12–19.
  33. 33. Andersen RM (1995) Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav 1995 36: 1–10.
  34. 34. Sahn D, Younger S, Genicot G (2003) The Demand for Health Care Services in Rural Tanzania. Oxford Bulletin Economics and Statistics. 65(2): 241–260.
  35. 35. Consejo Nacional de Población (CONAPO) (2012) Índice de marginación por localidad. Available: http://www.conapo.gob.mx/es/CONAPO/Indice_de_Marginacion_por_Localidad_2010. Accessed 2012 November 11.
  36. 36. Wirtz V, Serván-Mori E, Heredia-Pi I, Dreser A, Ávila-Burgos L (2013) Utilización y gasto en medicamentos en México: avances y retos en el acceso equitativo a medicamentos. Salud Publica Mex 2013 55: 00–00.
  37. 37. Rosenbaum P, Rubin D (1983) The central role of the propensity score in observational studies for causal effects. Biometrika 70 1: 41–55.
  38. 38. Dehejia R, Wahba S (2002) Propensity Score Matching Methods for Non-experimental Causal Studies. The Review of Economics and Statistics 2002 84(1): 151–161.
  39. 39. Caliendo M, Kopeinig S (2008) Some Practical Guidance for the Implementation of Propensity Score Matching, Journal of Economic Surveys, 22(1), pp. 31–72.
  40. 40. Cameron A, Trivedi P (2010) Microeconometrics using Stata A. College Station, Texas: Stata Press, 2010.
  41. 41. Stata Corp LP (2009) Stata/SE 13.0 for Windows XP 64 bits. College Station, Texas, USA: Stata Corp LP, 2009.
  42. 42. OMS (2012) Social Protection: Shared interests in vulnerability reduction and development. Social determinants of health sectorial briefing series, 4. WHO Document Production Services, Geneva, Switzerland, 2012.
  43. 43. Comisión Nacional de Protección Social en Salud (2013) Historia del Seguro Popular. Available: http://www.seguro-popular.salud.gob.mx/index.php?option=com_content&view=article&id=332&Itemid=326. Accessed 2013 August 28.
  44. 44. Gutiérrez JP, Hernández-Ávila M (2013) Cobertura de protección en salud y perfil de la población sin protección en México, 2000-2012. Salud Pública de México; 2013, 55 (sp 2): 83–96.