Health care utilization by adult Hispanic long-term survivors of hematopoietic stem cell transplantation

Report from the Bone Marrow Transplant Survivor Study

Authors


Abstract

BACKGROUND.

Long-term hematopoietic cell transplantation (HCT) survivors have a high prevalence of severe and chronic health conditions, placing significant demands on the healthcare system. The objective of the current study was to evaluate and compare the healthcare utilization by adult Hispanic and non-Hispanic white long-term survivors of HCT.

METHODS.

A mailed questionnaire was used to assess self-reported healthcare utilization in 3 domains: general contact with healthcare system, general physical examination outside cancer center (GPE), and cancer/HCT center visit. Eligible individuals had undergone HCT between 1974 and 1998, at age ≥21 years, and had survived ≥2 years after HCT.

RESULTS.

The cohort included 681 non-Hispanic white and 137 Hispanic survivors. The median age at HCT was 38.3 years, and the median length of follow-up was 6.6 years. Hispanic survivors had lower family income and education and were more likely to lack health insurance. The prevalence of GPE increased significantly over time among non-Hispanic whites (67% at 2-5 years to 76% at 11+ years) but remained unchanged among Hispanics (66% to 61%). Cancer/HCT center visits declined over time among both Hispanics and non-Hispanic whites, but a higher proportion of Hispanics reported cancer/HCT center visits at 11+ years after HCT (81% vs 54%).

CONCLUSIONS.

Compared with non-Hispanic whites, Hispanic survivors are less likely to establish contact with primary care providers years after HCT and to continue to receive care at cancer/HCT centers. Future studies of this population are needed to establish the factors responsible for this pattern of healthcare utilization. Cancer 2008. © 2008 American Cancer Society.

Ethnicity and race are important determinants of health outcome in various disease states. According to the 2005 census, Hispanics form the largest ethnic minority group in the US, constituting 14.4% of the entire population.1 Compared with non-Hispanic whites, Hispanics are considered to be a vulnerable population for adverse health outcomes in general2 and in the field of oncology in particular.3 Significant differences have been noted in the incidence of and mortality related to cancer in the Hispanic population compared with non-Hispanic whites.4, 5 The reasons for such differences include, but are not limited to, financial, cultural, and political barriers, as well as barriers within the healthcare system. Accessibility and continuity have been identified as key dimensions of primary care by the Institute of Medicine.6 Of all the minority racial and ethnic groups in the US, Hispanics are most likely to be uninsured.7–11 Furthermore, studies have shown that Hispanics are more likely to report lack of continuity of care or no usual source of care.12–14 However, to our knowledge, there is limited information regarding the patterns of healthcare utilization by Hispanic cancer survivors.15

Hematopoietic cell transplantation (HCT) has become an increasingly used therapeutic option for a variety of life-threatening malignant and nonmalignant conditions. High-dose chemotherapy, total body irradiation (TBI), and management of graft versus host disease (GVHD), with the attendant prolonged immune suppression, place the survivors of HCT at a particularly high risk of developing late complications necessitating an increased utilization of healthcare services.16–24

We have previously described the healthcare utilization patterns by a predominantly non-Hispanic white population of HCT survivors.25 However, the patterns of healthcare utilization by a comparable Hispanic HCT survivor population was not described in detail. The aim of the current study was to compare the patterns of healthcare utilization of adult Hispanic HCT survivors with those of non-Hispanic white HCT survivors, and to evaluate the risk factors associated with the lack of healthcare utilization in this population.

MATERIALS AND METHODS

Patients

The self-reported utilization of healthcare services by individuals enrolled in the Bone Marrow Transplant Survivor Study (BMTSS) was evaluated in this study. The BMTSS is a collaborative retrospective cohort study between City of Hope National Medical Center (COH) and the University of Minnesota (UMN). Survivors who met all the following criteria were eligible for participation in the BMTSS: 1) HCT performed at COH or UMN between 1974 and 1998; 2) survival of ≥2 years after HCT irrespective of current disease status; and 3) English- or Spanish-speaking. The current analysis was restricted to subjects: 1) aged ≥21 years at the time of HCT; 2) belonging to 1 of the following 2 ethnic/racial categories: non-Hispanic whites or Hispanics; and 3) alive at the time of study participation. The BMTSS was approved by the Human Subjects committee at the participating institutions, and informed consent was obtained according to the Declaration of Helsinki.

Between February 1999 and August 2004, a 255-item self-administered questionnaire was used to collect information from eligible participants regarding demographic characteristics, marital status, insurance coverage, education, income, employment, post-HCT complications, utilization of medical care, current health status (response options: excellent, very good, good, fair, or poor), and concerns for future health. This questionnaire was adapted from the questionnaire used in the Childhood Cancer Survivor Study with modifications to include items specific to an HCT population. A Spanish version of the questionnaire was used for monolingual Spanish-speaking participants. Details regarding primary diagnosis, HCT conditioning, chronic GVHD, and the prophylaxis and treatment of GVHD were obtained from the medical records, with the consent of the study participant.

Outcomes Measures

Self-reported healthcare utilization in the 2 years preceding the study was assessed. The 3 domains used to evaluate healthcare utilization included: 1) general contact with the healthcare system (general contact); 2) general physical examination outside a cancer/HCT center (GPE); and 3) cancer/HCT-related visit with the transplant team or medical visit at a cancer center (cancer/HCT center visit). These outcomes were not mutually exclusive. General or nonspecific medical contact was any contact with a physician, nurse, or other healthcare provider in the 2 years before the survey. A visit to a physician's office or a phone contact was also considered a general medical contact. Self-report of a general physical examination outside of a cancer/HCT center was considered as a GPE. All visits to see a physician at the cancer/HCT center, irrespective of the reason for the visit, were included as cancer/HCT center visits. No details regarding the content of the visit with any providers were collected.

Statistical Analysis

Outcome measures were analyzed for the entire cohort, and for Hispanic and non-Hispanic white HCT survivors separately. Comparisons between Hispanics and non-Hispanic whites were made using the Student t test for continuous variables and chi-square for categoric variables. Odds ratios (ORs) and 95% confidence intervals (95% CIs) for risk factors for absence of healthcare utilization were calculated by using unconditional logistic regression. Variables with a P value <.1 on univariate analysis were included in the stepwise logistic regression. The final multivariate regression model only included variables with P values <.05. The variables considered in the univariate analysis included sex, ethnicity (Hispanics vs non-Hispanic whites), age at time of HCT, age at study participation, educational status, household income, current health insurance, primary diagnosis, conditioning regimen (TBI vs non-TBI based), time since HCT, presence of chronic GVHD and its prophylaxis and treatment, type of transplantation (allogeneic vs autologous), risk of recurrence at HCT (standard vs high risk), current health status, and concerns for future health. Patients were considered at standard risk for recurrence if they were in first or second complete remission after acute leukemia or lymphoma, or first chronic phase of chronic myeloid leukemia. All other patients were placed into the high-risk category. The analysis was conducted for the entire cohort, and also stratified by type of transplant (autologous HCT and allogeneic [related and unrelated donor] HCT). Statistical analyses were performed using SAS software (version 9.1; SAS institute, Cary, NC).

RESULTS

Of the 1224 patients eligible for participation in this study, 1143 (93%) were successfully contacted and 818 (72%) agreed to participate. There were no differences between the 818 participants and the 406 nonparticipants in terms of sex (55% men vs 59% women; P = .16), type of transplant (autologous [49%] vs allogeneic [41%] vs matched unrelated donor [10%] vs 47%, 45%, and 8%, respectively; P = .35), primary diagnosis (chronic myeloid leukemia in 26.5%, acute myeloid leukemia in 22.5%, Hodgkin and non-Hodgkin lymphoma in 32.9%, acute lymphocytic leukemia in 6.0%, and other in 12.2% vs 24.6%, 20.0%, 32.8%, 8.4%, and 14.3%, respectively; P = .37), and risk of recurrence at HCT (61.6% vs 63.6%; P = .51). However, participants were older at HCT compared with nonparticipants (aged 39 years vs aged 36 years; P < .001). The age difference was predominantly observed in non-Hispanic whites (aged 40 years vs aged 37 years; P < .001). For Hispanic survivors, age at HCT was similar for participants and nonparticipants (aged 35 years vs aged 36 years; P = .72).

Self-reported race/ethnicity resulted in identification of 681 non-Hispanic whites and 137 Hispanics in this cohort (16.7%). Among the 137 Hispanics, 96 reported a reasonable understanding of written and spoken English, whereas 41 were monolingual Spanish-speaking.

The demographic characteristics of the entire cohort, by race/ethnicity and by language, are described in Table 1. Greater than half of the cohort was male, and the median age at HCT was 38 years. The median length of follow-up was 6.6 years (range, 2 years-24.4 years), and 56% of the cohort had been followed for >5 years. The Hispanic HCT survivors were significantly younger than non-Hispanic whites at the time of HCT (P = .02) and at study participation (P = .003). In addition, Hispanic survivors were significantly more likely to be uninsured (22.4% vs 4.6%; P < .001); to report a lower educational background (some high school or lower education: 37.5% vs 6%; P < .001); and to report household incomes below $20,000 (45.6% vs 8.8%; P < .001). The time from HCT to study participation was significantly longer for Hispanics when compared with non-Hispanic whites (mean follow-up time, 8.7 years vs 7.6 years; P = .01). Among the Hispanic survivors, the monolingual Spanish-speaking survivors were older at the time of HCT and at study participation, and were found to have lower education as well as household income when compared with their English-speaking counterparts.

Table 1. Demographic Characteristics of Hispanic and Non-Hispanic White Survivors of Hematopoietic Cell Transplantation
 Entire Cohort (N=818)Non-Hispanic White (n=681)Hispanic-White (n=137)
OverallP*English-speaking (n=96)Spanish-speaking (n=41)P
  • SD indicates standard deviation.

  • *

    P value for non-Hispanic versus Hispanic survivors.

  • P value for English-speaking versus Spanish-speaking Hispanic survivors.

Age at transplantation, y       
 Median (range)38.3 (21-68.6)39.0 (21.0-68.6)33.2 (21.1-62.4) 31.5 (21.2-62.4)37.3 (21.4-61.4) 
 Mean (SD)39.1 (10.7)39.9 (10.7)35.1 (10.2)<.00133.7 (10.0)38.2 (10.2).02
Age at study participation, y       
 Median (range)46.6 (23.2-73)47.4 (23.3-73.0)43.3 (25.3-67.4) 41.6 (25.3-66.8)47.4 (29.2-67.4) 
 Mean (SD)46.8 (9.8)47.5 (9.7)43.7 (9.9)<.00142.1 (9.8)47.6 (9.3).003
Follow up time, y       
 Median (range)6.6 (2-24.4)6.4 (2.0-24.4)7.9 (2.5-20.9) 7.3 (2.5-20.9)8.9 (4.8-19.2) 
 Mean (SD)7.8 (4.5)7.6 (4.4)8.7 (4.6).018.3 (4.9)9.4 (3.6).16
Male gender, no. (%)451 (55.1)367 (53.9)84 (61.3).1160 (62.5)24 (58.4).66
Education, no. (%)   <.001  <.001
 <High school92 (11.3)41 (6.0)51 (37.5) 14 (14.6)37 (92.5) 
 High school graduate and some college374 (45.9)326 (48.0)48 (35.3) 45 (46.9)3 (7.5) 
 College degree349 (42.8)312 (45.9)37 (27.2) 37 (38.5)0 (0) 
Household income, no. (%)   <.001  .008
 ≥$60,000/y358 (46.4)342 (52.9)16 (12.8) 16 (17.4)0 (0) 
 $20,000-$59,999/y300 (38.9)248 (38.3)52 (41.6) 39 (42.4)13 (39.4) 
 <$20,000/y114 (14.8)57 (8.8)57 (45.6) 37 (40.2)20 (60.6) 
Currently lacking health insurance, no. (%)61 (7.5)31 (4.6)30 (22.4)<.00118 (19.1)12 (30.0).17
Duration of follow-up, no. (%)   .12  .04
 2-5 y359 (43.9)309 (45.4)50 (36.5) 41 (42.7)9 (21.9) 
 6-10 y284 (34.7)233 (34.2)51 (37.2) 30 (31.2)21 (51.2) 
 ≥11 y175 (21.4)139 (20.4)36 (26.3) 25 (26.0)11 (26.8) 

Table 2 describes the clinical characteristics for the entire cohort by the 2 race/ethnicities of survivors (Hispanics vs non-Hispanic whites) and by the preferred spoken language. A significantly larger proportion of Hispanic survivors had undergone allogeneic HCT (62% vs 36%; P < .001) and received cyclosporine A (57% vs 35%; P < .001) as either prophylaxis or treatment for GVHD. Among the Hispanic survivors, English-speaking and Spanish-speaking Hispanic survivors were comparable in terms of primary diagnosis, risk of recurrence at HCT, and use of a TBI-based conditioning regimen. However, the monolingual Spanish-speaking Hispanic survivors were more likely to have been exposed to cyclosporine A for GVHD prophylaxis/treatment. A larger proportion of Hispanic survivors expressed lack of concern for future health compared with non-Hispanic white survivors (11% vs 4.5%; P = .002).

Table 2. Clinical Characteristics of Hispanic and Non-Hispanic White Hematopoietic Cell Transplantation Survivors
 Entire Cohort (N=818), No. (%)Non-Hispanic White (n=681), No. (%)Hispanic-White (n=137)
Overall, No. (%)P*English-speaking (n=96)Spanish-speaking (n=41)P
  • HL indicates Hodgkin lymphoma; NHL, non-Hodgkin lymphoma; ALL, acute lymphocytic leukemia; AML, acute myeloid leukemia; CML, chronic myeloid leukemia; HCT, hematopoietic cell transplantation; TBI, total body irradiation; GVHD, graft versus host disease; MUD, matched unrelated donor.

  • *

    P value for non-Hispanic versus Hispanic survivors.

  • P value for English-speaking versus Spanish-speaking Hispanic survivors.

Primary diagnosis   <.001  .59
 HL/NHL269 (32.9)243 (35.7)26 (19.0) 21 (21.9)5 (12.2) 
 ALL49 (6.0)31 (4.55)18 (13.1) 13 (13.5)5 (12.2) 
 AML183 (22.4)148 (21.7)35 (25.6) 25 (26.0)10 (24.4) 
 CML217 (26.5)179 (26.3)38 (27.7) 25 (26.0)13 (31.7) 
 Other100 (12.2)80 (11.7)20 (14.6) 12 (12.5)8 (19.5) 
High recurrence risk at HCT313 (38.4)266 (39.2)47 (34.3).2832 (33.3)15 (36.6).71
TBI conditioning regimen639 (78.4)533 (78.5)106 (77.9).8974 (77.9)32 (78.0).98
Chronic GVHD255 (31.2)204 (30.0)51 (37.2).1032 (33.3)19 (46.3).15
Cyclosporine A prophylaxis/treatment for GVHD316 (36.7)239 (35.1)77 (56.6)<.00148 (50.0)29 (72.5).02
Fair/poor current health172 (21.1)135 (19.9)37 (27.0).0624 (25.0)13 (31.7).42
Not concerned for future health45 (5.6)30 (4.5)15 (11.2).0028 (8.5)7 (17.5).13
HCT type   <.001  .13
 Allogeneic335 (41)250 (36.7)85 (62.0) 55 (57.3)30 (73.2) 
 MUD80 (9.8)75 (11.0)5 (3.7) 5 (5.2)0 
 Autologous403 (49.3)356 (52.3)47 (34.3) 36 (37.5)11 (26.8) 

Patterns of Healthcare Utilization

The overall healthcare utilization by ethnicity and spoken language, and stratified by type of transplantation, is shown in Table 3. Hispanic survivors were significantly less likely to report general contact with the healthcare system (93% vs 98%; P = .001) in the 2 years preceding the study, when compared with the non-Hispanic whites. Both groups reported comparable prevalence of GPE (66% vs 71%; P = .23). However, Hispanic survivors were significantly more likely to report a visit to the Cancer/HCT center than non-Hispanic whites (88% vs 81%; P = .03). Among Hispanic survivors, the monolingual Spanish-speaking Hispanics were significantly less likely to report GPE (49% vs 73%; P = .006) compared with English-speaking Hispanics.

Table 3. Healthcare Utilization by Hispanic and Non-Hispanic White Survivors of Hematopoietic Cell Transplantation
 Entire Cohort (N=818), No. (%)Non-Hispanic White (n=681), No. (%)Hispanic-White (n=137)P
Overall, No. (%)P*English-speaking (n=96)Spanish-speaking (n=41)
  • HCT indicates hematopoietic cell transplantation; GPE, general physical examination outside cancer center; MUD, matched unrelated donor.

  • *

    P value for non-Hispanic versus Hispanic survivors.

  • P value for English-speaking versus Spanish-speaking Hispanic survivors.

Any HCT       
 General contact797 (97.4)669 (98.2)128 (93.4).00189 (92.7)39 (95.1).72
 GPE573 (70.0)483 (70.9)90 (65.7).2370 (72.9)20 (48.8).006
 Cancer/HCT670 (81.9)549 (80.6)121 (88.3).0384 (87.5)37 (90.2).78
Autologous-HCT only       
 General contact392 (97.3)347 (97.5)45 (95.4).4934 (94.4)11 (100.0)1.00
 GPE295 (73.2)263 (73.9)32 (68.1).4026 (72.2)6 (54.6).27
 Cancer/HCT333 (82.6)291 (81.7)42 (89.4).2031 (86.1)11 (100).32
Allogeneic/MUD HCT Only       
 General contact405 (97.6)322 (99.1)83 (92.2).00155 (91.7)28 (93.3)1.00
 GPE278 (67.0)220 (67.7)58 (64.4).5644 (73.3)14 (46.7).01
 Cancer/HCT337 (81.2)258 (79.4)79 (87.8).0753 (88.3)26 (86.7).82

When stratified by type of transplant, a similar pattern was observed in both groups. For the entire cohort, Hispanic survivors were more likely to report cancer/HCT center visits compared with non-Hispanic whites. The differences by ethnicity were not found to be statistically significant among survivors of autologous HCT. For allogeneic HCT survivors, Hispanics were less likely to report general contact (P = .001) and more likely to report cancer center/HCT-related visits (P = .07) when compared with non-Hispanic whites. Furthermore, the monolingual Spanish-speaking Hispanics were less likely to report GPE compared with English-speaking Hispanics in both transplant groups, but the difference was significant in survivors of allogeneic HCT only (47% vs 73%; P = .01).

Healthcare utilization, as measured by self-reported general medical contact, GPE, and visits to Cancer/HCT center, as a function of time after HCT, is shown in Figure 1.

Figure 1.

Healthcare utilization is depicted as a function of time from transplantation in adult Hispanic and non-Hispanic white survivors of hematopoietic cell transplantation (HCT). GPE indicates general physical examination outside cancer center.

Healthcare utilization reported by non-Hispanic whites

The prevalence of general medical contact remained high and did not change significantly over time (99% at 2-5 years, 98% at 6-10 years, and 97% at 11+ years after HCT; P for trend = .26). There was a significant increase in the prevalence of GPE from 67% at 2 to 5 years, to 73% at 6 to 10 years and 76% at 11+ years after HCT (P for trend = .05). Conversely, cancer/HCT center visits declined significantly from 94% at 2 to 5 years to 79% at 6 to 10 years and 54% at 11+ years after HCT (P for trend <.001).

Healthcare utilization reported by Hispanics

Greater than 90% of Hispanic survivors continued to report general medical contact up to 11+ years from transplantation, and this did not change significantly over time (96% at 2-5 years, 92% at 6-10 years, and 92% at 11+ years after HCT; P for trend = .40). The prevalence of GPE also did not change significantly over time (66% at 2-5 years and 69% at 6-10 years vs 61% at 11+ years after HCT; P for trend = .68). However, the cancer/HCT center visits declined over time from 96% at 2 to 5 years to 86% at 6 to 10 years and 81% at 11+ years after HCT (P for trend = .03).

Comparison of healthcare utilization between non-Hispanic whites and Hispanics

The prevalence of general medical contact remained high over time and did not differ between Hispanic and non-Hispanic survivors (P for follow-up time and race interaction = .95). Although there was a significant increase in the prevalence of self-reported GPE over time in non-Hispanic whites, and no significant change was observed in Hispanic survivors, the difference was not statistically significant (P for interaction = .23). In both groups, cancer/HCT center visits declined over time. However, this decline in cancer/HCT center visits did not appear to differ significantly between the 2 groups over time (P for interaction = .23).

Risk Factors for Lack of Healthcare Utilization

Entire cohort

The results of multivariate analysis for risk factors associated with lack of healthcare utilization for the entire cohort are shown in Table 4.

Table 4. Risk Factors for the Absence of Healthcare Utilization in Entire Cohort of Hispanic and Non-Hispanic Survivors of Hematopoietic Cell Transplantation
Risk Factors*Entire Cohort (N=818), RR (95% CI)
General ContactGPECancer/HCT
  • RR indicates relative risk; 95% CI, 95% confidence interval; GPE, general physical examination outside cancer center; HCT, hematopoietic cell transplantation; CSA, cyclosporine A.

  • *

    Only risk factors found to be statistically significantly associated with the absence of healthcare utilization on multivariate analysis are listed in the table. No association was found in either group with other factors such as household income, type of transplantation, recurrence risk, and exposure to total body irradiation or methotrexate.

Race   
 Non-Hispanic white  1.00
 Hispanic  0.42 (0.22-0.80)
Language   
 English 1.00 
 Spanish 2.56 (1.35-4.87) 
Gender   
 Male1.00  
 Female0.23 (0.07-0.78)  
Age at time of HCT, y   
 18-45  1.00
 >45  0.54 (0.31-0.92)
Age at study participation, y   
 <40 1.00 
 40-44 0.68 (0.43-1.07) 
 45-54 0.79 (0.54-1.16) 
 ≥55 0.48 (0.30-0.77) 
Follow-up since HCT, y   
 2-5  1.00
 6-10  3.39 (1.96-5.87)
 ≥11  8.29 (4.64-14.81)
Concerns of future health   
 Concerned  1.00
 Not concerned  3.97 (1.95-8.09)
Current health   
 Good 1.00 
 Fair/poor 1.96 (1.38-2.80) 
Current insurance   
 Insured1.00  
 Uninsured3.82 (1.33-1.02)  
Exposure to CSA   
 No exposure  1.00
 Exposed  0.44 (0.28-0.71)
General medical contact

Uninsured individuals were significantly less likely to have general medical contact than those who had medical insurance (OR of 3.82; 95% CI, 1.33-11.02). Compared with men, female HCT survivors were more likely to report general medical contact (OR [lack of reporting] of 0.23; 95% CI, 0.07-0.78).

GPE

Individuals who were monolingual Spanish-speaking (OR of 2.56; 95% CI, 1.35-4.87), and those reporting fair/poor current health (OR of 1.96; 95% CI, 1.38-2.80) were significantly more likely to report lack of GPE. Individuals who were older at the time of study participation were less likely to report lack of GPE (P for trend = .007).

Cancer/HCT center visit

Hispanic race (OR [for lack of reporting a visit] of 0.42; 95% CI, 0.22-0.80), age >45 years at HCT (OR [lack of reporting a visit] of 0.54; 95% CI, 0.31-0.92), and exposure to cyclosporine A (OR [lack of reporting a visit] of 0.44; 95% CI, 0.28-0.71) were found to be significantly associated with reporting a cancer/HCT center visit. Increasing follow-up time (OR of 8.29; 95% CI, 4.64-14.81) and lack of concern for future health (OR of 3.97; 95% CI, 1.95-8.09) were associated with lack of cancer/HCT center visit. When stratified by type of transplant, in both autologous and allogeneic HCT groups, Hispanic race was associated with reporting a cancer/HCT center visit (OR [lack of reporting a visit] of 0.41; 95% CI, 0.14-1.19; and OR of 0.41; 95% CI, 0.18-0.92). The association is similar in magnitude but only statistically significant in the allogeneic HCT group.

Multivariate analyses of risk factors for lack of healthcare utilization among non-Hispanic whites and Hispanics are shown in Table 5.

Table 5. Risk Factors for the Absence of Healthcare Utilization in Non-Hispanic White and Hispanic Survivors of Hematopoietic Cell Transplantation
Risk Factors*Non-Hispanic White, RR (95% CI)Hispanic White, RR (95% CI)
GPECancer/HCTGeneral ContactGPECancer/HCT
  • RR indicates relative risk; 95% CI, 95% confidence interval; GPE, general physical examination outside cancer center; HCT, hematopoietic cell transplantation; CSA, cyclosporine A.

  • *

    Only risk factors found to be statistically significantly associated with the absence of healthcare utilization on multivariate analysis are listed in the table.

Follow-up, y     
 2-5 1.00  1.00
 6-10 3.76 (2.11-6.71)  3.82 (0.75-19.37)
 ≥11 11.87 (6.56-21.47)  5.79 (1.13-29.80)
Age at study participation, y     
 <401.00    
 40-440.68 (0.41-1.14)    
 45-540.80 (0.52-1.23)    
 ≥550.43 (0.26-0.73)    
Education     
 High school or less   1.00 
 High school and college   0.34 (0.16-0.71) 
Concerns for future health     
 Concerned 1.001.00  
 Not concerned 5.09 (2.22-11.67)4.71 (1.04-21.28)  
Current health status     
 Good1.001.00   
 Fair/poor2.10 (1.41-3.12)0.47 (0.24-0.89)   
Exposure to CSA     
 No exposure 1.00   
 Exposed 0.54 (0.33-0.89)   

Non-Hispanic whites

GPE

Fair or poor current health (OR of 2.10; 95% CI, 1.41-3.12) was significantly associated with lack of GPE. Older age at study participation was significantly associated with reporting of GPE (P for trend = .007).

Cancer/HCT center visit

Lack of concern for future health (OR of 5.09; 95% CI, 2.22-11.67) and longer follow-up (OR of 6.34 at 11+ years; 95% CI, 3.46-11.6) were significantly associated with lack of HCT/cancer center visit. Exposure to cyclosporine A (OR [lack of cancer/HCT visit] of 0.54; 95% CI, 0.33-0.89) was significantly associated with reporting cancer/HCT center visit.

Hispanics

General medical contact

Lack of concern for future health was significantly associated with reporting lack of general medical contact (OR of 4.71; 95% CI, 1.04-21.28).

GPE

Individuals with high school or college education were more likely to report a GPE (OR [lack of GPE] of 0.34; 95% CI, 0.16-0.71).

Cancer/HCT center visit

Increasing time of follow-up was the only independent risk factor found to be significantly associated with the lack of cancer/HCT center visit for Hispanic survivors (OR of 5.79 at 11+ years; 95% CI, 1.13-29.8).

DISCUSSION

This study of a large cohort of HCT survivors demonstrated that the prevalence of general medical contact is high even up to 11+ years after HCT. For the entire cohort of HCT survivors, individuals uninsured at the time of study participation and male survivors were significantly less likely to have any medical contact. Monolingual Spanish-speaking Hispanics as well as those reporting fair/poor current health and those younger at study participation were less likely to report GPE. Overall, Hispanics were more likely to report cancer/HCT center visits, as were HCT survivors aged >45 years, and those who had received cyclosporine for prophylaxis or treatment of GVHD. The likelihood of cancer/HCT center visit decreased with time from HCT and among those who reported lack of concern for future health.

We described the healthcare utilization by HCT survivors in a predominantly white population in a previous report,25 demonstrating that the prevalence of GPE increased with time, whereas that of cancer/HCT center visits decreased. However, a detailed examination of the healthcare utilization pattern reported by Hispanic HCT survivors and its comparison with the non-Hispanic white survivors were not described, and are the subject of the current report. In the current report, although >90% of both Hispanic and non-Hispanic white survivors reported medical contact 11+ years after HCT, significant differences were noted in the pattern of healthcare utilization between the 2 populations. The prevalence of GPE increased significantly with increasing time from transplantation among non-Hispanic whites, whereas it remained largely unchanged among Hispanics. Self-reported visit to cancer/HCT centers decreased over time in both groups.

Differences in healthcare utilization by Hispanic and non-Hispanic whites have been reported in childhood cancer survivors. Castellino et al compared the long-term outcomes, healthcare utilization, and health-related behaviors of minority adult childhood cancer survivors to those of white survivors in the Childhood Cancer Survivor Study (CCSS).15 Regardless of the socioeconomic status, black female and Hispanic male survivors were found to have significantly less general contact with the medical system. Hispanic survivors were more likely to report a visit to a cancer center than non-Hispanic white survivors. Male sex, lack of health insurance, lack of concern for future health, and age >30 years at the time of the study were associated with lack of reporting GPE, cancer-related visits, or a cancer center visit among childhood cancer survivors in another study from the CCSS cohort,26 a pattern of healthcare utilization that is quite similar to that identified in the current study.

Increased utilization of hospital-based care for cancer therapy among patients who are uninsured or insured with Medicaid has been shown in other studies.27 A cross-sectional study using the National Ambulatory Medical Care Survey and National Hospital Ambulatory Care Survey for care of cancer patients showed that patients with Medicaid were more likely to visit hospital clinics rather than private offices for treatment as compared with those who were privately insured.27 Our study demonstrates that Hispanics were more likely to use the cancer/HCT center long-term after adjustment for current insurance status and other relevant medical factors that could have determined patterns of utilization, indicating that other sociocultural variables that remain unmeasured in this study could have influenced this pattern of utilization.

Health insurance and cost barriers to care among cancer survivors have been studied using the 1998 and 2000 National Health Interview Survey.28 Uninsured and publicly insured survivors were more likely to delay or miss care because of cost. Overall, 68% of uninsured survivors reported delaying or missing needed care. In the current study, a significantly higher number of Hispanic survivors reported a lack of health insurance when compared with non-Hispanic whites, and a lack of health insurance was noted to significantly decrease the general medical contact in the cohort of HCT survivors.

Hispanics experience substantial barriers to primary care. Previous studies have shown that Hispanics are more likely to report lack of continuity of care or no usual source of care.12, 13 Proficiency in English language has been postulated as 1 of the possible mechanisms for disparities in quality of primary care between Hispanic and non-Hispanic whites. Analysis of data on insured Latinos from the National Latino and Asian American Study, a nationally representative household survey taken during 2002 to 2003, showed that low English language proficiency was associated with worse reports of quality of primary care.29 Language proficiency has also been associated with lack of utilization of preventive healthcare by Hispanics.30 Other studies have associated low English proficiency with less timeliness of care as well as poor communication with providers and less helpful staff.31

A cross-sectional analysis of the Community Tracking Survey (1996-1997) that studied adults ages 18 to 64 years with private or Medicaid health insurance found that the pattern of healthcare utilization for English-speaking Hispanic patients was not significantly different from non-Hispanic whites.7 In contrast, the Spanish-speaking Hispanic patients were significantly less likely to have had a physician visit, mental health visit, or influenza vaccine as compared with the non-Hispanic whites.

Our study also points to English proficiency as a determinant of healthcare utilization. Spanish-speaking survivors were 2 times more likely not to have a GPE as compared with English-speaking survivors. Among Hispanics, higher educational status was associated with greater likelihood of having undergone a GPE, most likely indicating that a higher educational status was more likely to be associated with proficiency in the English language.

The results of the current study demonstrated that with increasing time from HCT, the survivors were less likely to report a cancer/HCT center visit for their medical care, a finding that was true for both Hispanic and non-Hispanic whites. However, Hispanics reported a higher utilization of the cancer/HCT center compared with non-Hispanic whites. The reason for this difference is not clear. Similar patterns were observed for both autologous and allogeneic HCT survivors, and the prevalence of chronic GVHD did not differ between Hispanic and non-Hispanic white allogeneic survivors; therefore, the burden of post-HCT morbidity was unlikely to contribute to these observed differences in healthcare utilization. We speculate that prolonged utilization of the cancer/HCT center could be a reflection of lack of a primary care provider to provide general care.

To our knowledge, this is the first study evaluating healthcare utilization in Hispanic survivors of HCT. However, the results of the study need to be interpreted in the context of certain limitations. The total number of Hispanic survivors is relatively small, and the number of monolingual Spanish-speaking Hispanic survivors even smaller. Hence, for the most part, both English-speaking and Spanish-speaking Hispanic survivors were evaluated as a single population for the risk factors associated with lack of healthcare utilization. The inclusion of English-speaking Hispanics is likely to decrease the observed differences in the patterns of healthcare utilization between Hispanics and non-Hispanic whites. Nevertheless, the study was able to demonstrate significant differences between the Hispanic and non-Hispanic white survivors. Another limitation stems from the finding that healthcare utilization was determined by self-report and was not verified by healthcare providers. The reasons for the visits to cancer/HCT centers were not solicited. These limitations notwithstanding, the results of this study demonstrate that Hispanic survivors of HCT continue to receive care at cancer/HCT centers several years after HCT. Future studies are needed to further investigate the factors responsible for such differences among Hispanic and non-Hispanic white populations.

Ancillary