ACCESS INEQUALITIES BETWEEN PRIVATE AND PUBLIC INSURANCE AMONG CANCER PATIENTS:RESULTS OF A NATIONAL SURVEY IN BRAZILFlores NM1, Minowa E2, Lee LK1, Vietri J3, Pomerantz D3

PCN52

Poster Presented at the ISPOR 5th Latin America Conference 6 -8 September 2015 Santiago, Chile ©Copyright 2015 Evidências Credibilidade Científica, A Kantar Health Company Alameda Lorena, 131, Conj. 115 e 117, Jardim Paulista São Paulo, SP - 01424-001 Brazil + 55-11-3884-0942

www.evidencias.com.br www.kantarhealth.com

1Kantar Health, Foster City, CA, USA; 2Evidências _ Kantar Health, São Paulo, Brazil; 3Kantar Health, Horsham, PA, USA

Further research is needed to understand why inequality of access is not reflected in HRQoLdecrements in this population.

Along with the differences observed between insurance type and mood disorders and sleepconditions, additional investigation is required to understand the relationship between socialinequalities and its impact on prevalence.

Overall, the present study showed an urgent need to develop health policy strategies in Brazil in order to improve equality in healthcare access among cancer patients utilizing the public system. Also, drug access policy should be addressed.

CONCLUSION

•

•

•

Self-reported cannot be verified by patients’ medical charts or other objective data.Cross-sectional data do not allow for causal explanations to be made. Although a number of respondent characteristics were controlled for, there may be additional variables that were not controlled for, which may have affected the results.

LIMITATIONS

•

Lower healthcare resource use among those on the PuHS indicates disparities in healthcare access among cancer patients in Brazil. Results are consistent with prior studies evaluating overall healthcare access. 2,9

The present study showed differences on healthcare use measured by number of traditional HCP visits and number of ER visits and hospitalizations, however, further investigation is required to evaluate the inequalities on drug access.

HRQoL was not associated with insurance type, which suggests that decrements in quality of life in cancer patients may be more influenced by disease burden than insurance and quality of treatment.

Although social inequalities are often associated with higher disease prevalence, the present study showed that anxiety and depression were more common among those with private insurance. This result may be associated with greater self-awareness of disease signs and symptoms, which may be related to better healthcare access such as receiving diagnosis and obtaining guidance from healthcare professionals.2

DISCUSSION

•

•

•

•

METHODOLOGY

•

•

•

•

•

•

•

•

Data came from the 2011, 2012, & 2014 Brazil National Health and Wellness Survey (NHWS; N=36,000), a self-administered, internet-based general health questionnaire from a sample of adults (aged 18+), supplemented by offline recruitment to reach the elderly (aged 65+) that may not have Internet access.

Stratified random sampling was used to ensure the demographic composition (with respect to age and gender) is representative of the adult population of Brazil as reported by theInternational Data Base of the U.S. Census Bureau and Organization for Economic Cooperation and Development.

The sample included 1,122 respondents who reported a cancer diagnosis.

Independent Variable: respondents were categorized as to insurance type into public insurance (n=448) or private insurance (n=674).

Demographics, health characteristics, cancer type, mood disorders and sleep conditions were all examined.

Health-Related Quality of Life: measured via the revised Medical Outcomes Study Short Form Health Survey (SF-12v2 for 2011 NHWS; SF36v2 for 2012 & 2014 NHWS) including 6-8: MentalComponent Summary Score (MCS), Physical Component Summary Score (PCS), Health Utilities (SF-6D)

Healthcare resource use: Number of visits for the past 6 months to traditional healthcare provider (HCP), Emergency Room and Hospitalizations.

One-way ANOVAs (for continuous variables) and Chi-square tests (for categorical variables) were used to examine differences in demographics, health characteristics, and mood/sleep disorders across type of insurance.

Generalized linear models were then used to predict health outcomes from insurance type controlling for demographics and health characteristics.

Any two-sided p-value <0.05 was considered statistically significant.

Outcome Measures

•

Data Source

Sample

Statistical Analyses

Measures

ReferencesBrasil. Tribunal de Contas da União. Relatório de auditoria operacional Política Nacional de Atenção Oncológica / Tribunal de Contas da União;Relator Ministro José Jorge. – Brasília: TCU, Secretaria de Fiscalização e Avaliação de Programas de Governo, 2011.132 p.Barros MBA, César CLG, Carandina L, G.D. T. Desigualdades sociais na prevalência de doenças crônicas no Brasil, PNAD-2003. . Ciênc. saúde coletiva. 2006;11(4):9110926.Barros MBA, Francisco PMSB, Zanchetta LM, César CLG. Tendências das desigualdades sociais e demográficas na prevalência de doenças crônicas no Brasil, PNAD: 2003- 2008. . Ciênc. saúde coletiva. 2011;16(9):3755-3768.Instituto Nacional do Câncer. Estimativa 2014: Incidência de Câncer no Brasil / Instituto Nacional de Câncer José Alencar Gomes da Silva, Coordenação de Preven-ção e Vigilância. Rio de Janeiro: INCA, 2014.Fórum Nacional de Políticas de Saúde em Oncologia, 5, 2015, Brasilia, Brazil. Lecture: Acesso ao tratamento no SUS Available at: http://www.oncoguia.org.br/onco-guia-material/redes-de-atencao-a-saude-prioridades-sandro-martins/129/36/ Ware J, Jr., Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care. 1996;34(3):220-233.Ware JE, Jr.,, Kosinski M, Turner-Bowker DM, Gandek B. User’s Manual for the SF-12v2® Health Survey (With a Supplement Documenting SF-12® HealthSurvey). Lincoln, RI: QualityMetric Incorporated and Health Assessment Lab2002.Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30(6):473-483.Machado M, DiBonaventura MC, Pomerantz D. The impact of private verus public insurance on health status, work productivity and health care utilization ofr indi-viduals residing in Brazil. . ISPOR 18th Annual International Meeting. 2013.

Figure 1. Mood Disorders and Sleep Conditions by Insurance Type

Figure 2a. Adjusted Mean Number of Traditional HCP Visits in the Past6 Months By Insurance Type

Figure 2b. Adjusted Mean Number of ER Visits and Hospitalizations in thePast 6 Months by Insurance Type

Cancer represents a burden of disease with an increasing incidence worldwide. Approximately 576,000 new cases of cancer are expected each year in Brazil.1

About 70% of the population in Brazil is covered only by the public health system (PuHS). Previous studies showed inequalities for overall health access with higher prevalence of office visits and hospitalizations for those with private insurance (PHS).2,3

Little is known regarding access inequalities between health systems for patients with cancer. According to the Ministry of Health, there are inadequate numbers of specialized oncology public institutions to properly assist the cancer population.4,5 A total of 307 specialized oncology institutions should be available to patients in PuHS; however, there are only 283 suchestablishments in the country, indicating a lack of sufficient resources.4

It is important to understand how different types of insurance impact health outcomes in order to prioritize problems and raise awareness for health policies. Quality of life is essential to oncology patients; however, it is often undervalued by the healthcare system. Also, anxiety and depression, highly prevalent among patients with cancer, are related to both the disease and its treatment.

Our goal was to evaluate the real world impact of insurance type on health-related quality of life (HRQoL), mood and sleep-related outcomes, and healthcare resource use for patients with cancer.

INTRODUCTION

•

•

•

•

•

Mean SD Mean SDAge 47.41 17.36 46.00 15.25 .153Charlson Comorbidity Index (CCI) 1.72 2.02 2.04 2.62 .027

Column % Count Column % Count p-value*Sex: Female 55.8% 250 54.0% 364 .554Ethnicity: White 67.2% 301 76.6% 516 <.001Education: Degree Completed or Higher 30.1% 135 57.6% 388 <.001 <R$ 1000 15.0% 67 3.3% 22 <.001 R$ 1001 to R$ 6500 43.5% 195 38.4% 259 R$ 6501+ 27.0% 121 49.7% 335 Decline to Answer 14.5% 65 8.6% 58 Underweight (<18.5) 3.8% 17 2.1% 14 .045 Normal Weight (18.5-24.99) 41.7% 187 37.5% 253

Overweight (25-29.99) 34.8% 156 35.3% 238 Obese (>30) 18.1% 81 24.2% 163 Unknown 1.6% 7 0.9% 6Smoking Current Smoker 23.4% 105 19.0% 128 .191Status Former Smoker 32.1% 144 33.2% 224 Never Smoker 44.4% 199 47.8% 322Drink Alcohol 53.3% 239 68.4% 461 <.001Exercise 20+ minutes 1+ times in past 51.8% 232 55.6% 375 .205

Insurance TypePublic (n=448) Private (n=674) p-value*Characteristics

Table 1. Demographics and Health Characteristics by Insurance Type

Note: SD=standard deviation; BMI=Body mass index

Note: n.s.=not significant; *=p<.05; +=p<.10

Column % Count Row % Count Row % CountAny Tumor 35.2% 395 36.5% 144 63.5% 251Skin Cancer 17.8% 200 36.0% 72 64.0% 128Breast Cancer 11.2% 126 42.9% 54 57.1% 72Other Cancer 10.3% 116 35.3% 41 64.7% 75Prostate Cancer 8.6% 97 40.2% 39 59.8% 58Uterine Cancer 7.1% 80 61.3% 49 38.8% 31Colorectal Cancer 5.4% 61 32.8% 20 67.2% 41Lymphoma 5.3% 60 35.0% 21 65.0% 39Metastatic Solid Tumor 3.8% 43 37.2% 16 62.8% 27Ovarian Cancer 3.3% 37 45.9% 17 54.1% 20Leukemia 2.8% 31 51.6% 16 48.4% 15Non-Small Cell Lung Cancer 1.6% 18 33.3% 6 66.7% 12Small Cell Lung Cancer 1.5% 17 23.5% 4 76.5% 13Cervical Cancer 1.4% 16 31.3% 5 68.8% 11

Insurance TypeTotal(n=1,122 ) Public (n=448) Private (n=674)Type of Cancer

Table 2. Types of Cancer by Insurance Type

Note: Cancer types are not mutually exclusive as respondents could report having more than one type of cancer. Therefore, total percentage of reported cancer diagnoses exceeds 100%.

Note: Comparison between Public and Private Insurance was significant at p<.001 level. Barsrepresent standard errors. Covariates Included: age, gender, ethnicity, education, income, BMI,alcohol use, cigarette use, and exercise activity.

Note: Both comparisons between Public and Private Insurance were significant at p<.01 level. Bars represent standard errors. Covariates Included: age, gender, ethnicity, education, income, BMI,alcohol use, cigarette use, and exercise activity.

Private Insurance (n=674) p-value*

Note: SE=Standard Error. Covariates included age, gender, ethnicity, education, income, BMI,alcohol use, cigarette use, and exercise activity.

Estimated Mean SE Estimated

Mean SE

Mental Component Summary 44.01 0.54 44.56 0.43 .439Physical Component Summary 47.97 0.43 47.44 0.35 .356Health Utilities (SF-6D) 0.673 0.006 0.672 0.005 .896

Public Insurance(n=448)

Private Insurance(n=674) Outcome Measures

Table 3. Adjusted Health-related Quality of Life by Insurance Type

•

RESULTS

Those with private insurance were more likely to be white (76.6%), have completed their degree (57.6%), have annual household income ≥R$6501 (49.7%), be obese (24.2%), and drink alcohol (68.4%; see Table 1).

Sample Demographics & Health Characteristics

Overall, solid tumors (e.g. prostate cancer, uterine cancer) were more prevalent than hemato-logic neoplasia (e.g. lymphoma, leukemia). Most common solid tumors in Brazil regardless of insurance included skin cancer, breast cancer, uterine cancer and prostate cancer.

Rates of insurance type varied according to cancer type such as more patients with breast cancer had private insurance than public insurance whereas the opposite was true for patients with uterine cancer (see Table 2).

•

•

Cancer Type

Those with private insurance reported experiencing generalized anxiety disorder and any sleepcondition in the past 12 months significantly more than those with public insurance (see Figure 1).

Additionally, those with private insurance reported experiencing any anxiety disorder, sleep difficulties, and insomnia marginally more than those with public insurance (see Figure 1).

Health-related Quality of Life: after controlling for covariates, there were no significantdifferences for HRQoL by insurance type for MCS, PCS, or SF-6D (see Table 3).

Healthcare Resource Use: Those who had private insurance had greater healthcare resource use as measured by number of traditional HCP visits (see Figure 2a) and number of ER visits and hospitalizations (see Figure 2b) than those with public insurance.

•

•

•

•

Mood Disorders and Sleep Conditions

Health Outcomes

Public Insurance Private Insurance

5,4%

0

2

4

6

8

10

12

Mea

n N

umbe

r of V

isits

Tradi�onal Healthcare Provider

6.61

9.76

Public Insurance Private Insurance

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Mea

n N

umbe

r of V

isits

Emergency Room Hospitaliza�ons

0.53

1.02

0.50

0.30

Public Insurance Private Insurance

0%

10%

20%

30%

40%

50%

60%

Sam

ple

Perc

enta

ge

Mood and Sleep Disorders

Depression GeneralizedAnxiety Disorder

Any AnxietyDisorder

Insomnia Sleep Difficul�es Any SleepCondi�on

26.6%

n.s.

5.1%

43.5%

32.8%29.9%27.7%

9.2%

48.5%

37.8%35.3%

42.4%

49.1%

*

*

**

*

p-value

Annual Household Income

SmokingStatus

BMI Category

1.

2.3.

4.

5.

6.7.

8.9.