Abstract
Background:
Early treatment of hepatocellular carcinoma (HCC) is associated with improved survival but many patients with HCC do not receive therapy. We aimed to examine factors associated with HCC treatment and survival among incident HCC cases in a statewide cancer registry.
Patients and Methods:
All HCC cases from 2003–2013 were identified from the North Carolina cancer registry. These cases were linked to insurance claims from Medicare, Medicaid and large private insurers. We examined the association between pre-specified covariates with more advanced HCC stage at diagnosis (i.e. multifocal cancer), visit at a liver transplant center and provision of HCC treatment by multivariate logistic regression. A Cox proportional hazards model was developed to assess the association between these factors and survival.
Results:
Of 1,809 patients with HCC, 53% were seen at a transplant center <90 days from diagnosis, with lower odds in blacks [adjusted odds ratio (aOR 0.54, 95% CI 0.39,0.74)], Medicare insurance (aOR 0.35, 95% CI 0.21,0.59), Medicaid insurance (aOR 0.46, 95% CI 0.28,0.77), and rural patients; odds of transplant center visits were higher with pre-diagnosis AFP screening (aOR 1.74, 95% CI 1.35,2.23) and GI care (aOR 1.66, 95% CI 1.27,2.18). Treatment was more likely with pre-diagnosis GI care (aOR 1.68, 95% CI 0.98,2.86) and transplant center visit (aOR 2.42, 95% CI 1.74,3.36). Survival was strongly associated with age, stage, cirrhosis complications, and receipt of HCC treatment. Medicare (aHR 1.58, 95% CI 1.20,2.09) and Medicaid (aHR 1.55, 95% CI 1.17, 2.05) recipients had shorter survival than privately insured patients.
Discussion:
In this population based cohort of patients with HCC, Medicare/Medicaid insurance, rural residence and black race were associated with lower provision of HCC treatment and poorer survival. Efforts should be made to improve access to care for these vulnerable populations.
Keywords: Hepatocellular carcinoma, disparities, Medicare, Medicaid, quality of care
Introduction
Hepatocellular carcinoma (HCC) incidence and mortality are on the rise in the United States.1,2 The prognosis of HCC is poor, in part because potentially curative treatments, including surgical resection, ablation, and transplantation, are feasible in a minority patients with early stage HCC.3 In the absence of curative therapy, liver-directed locoregional therapies (LRTs) and drug therapy prolong survival, yet up to half of all patients with HCC never receive any cancer-directed therapy.4–7
Despite accumulating evidence that provision of timely HCC treatment decreases cancer related mortality,8 treatment rates remain low and the reasons for non-receipt of therapy in patients with HCC remains unclear. Medical comorbidities, decompensated liver disease and advanced HCC often precludes specific HCC-directed therapies. Other potential factors have been associated with low provision of HCC therapy including older patient age, patient insurance status and care at a low volume center.9–11 In addition, racial and ethnic disparities in the receipt of HCC treatment and survival have been well documented.12–16 Lastly, subspecialist consultation has been associated with improved treatment outcomes in the US Veterans Affairs system.17 This evidence suggests that both patient- and facility-level factors likely influence provision of HCC treatment. However, relatively few data sources allow for the simultaneous investigation of these many potential variables in a population-based sample.
We therefore assessed the potential predictors of advanced cancer, care at a transplant center, provision of treatment and mortality among patients from a population-based state registry linked to insurance claims data.
Materials and Methods
This work was approved by the Biomedical IRB at the University of North Carolina (#12–1828).
Patients
The cohort is comprised of patients diagnosed with HCC between 2003–2013, identified from the North Carolina Central Cancer Registry (NCCCR) by International Classification of Diseases (ICD)-10 C22.0 and histology codes 8170–8175, 8180. These NCCCR cases were linked to claims from Medicare, NC Medicaid, and large private insurers in NC by the UNC Lineberger Cancer Information & Population Health Resource.18 This subset consisted of patients with continuous healthcare enrollment in any health plan for 12 months preceding and 12 months following diagnosis (or death). Patients participating in a Medicare HMO/advantage plan were excluded as claims are not required to be reported to Medicare.
To evaluate factors associated with treatment and survival, only patients surviving the 90 day exposure window after diagnosis were included in multivariable models, thereby including only patients who may have been eligible for treatment.
Covariates
Patient age, sex, marital status, race, county and zip code of residence, and insurance status at diagnosis were derived from NCCCR demographics file. We used the NCCCR collaborative staging extension variables of number of tumors and presence/absence of vascular invasion to group cancers into clinically meaningful categories (single or multiple, with or without vascular invasion, and extrahepatic disease). Multifocal cancer was defined as the presence of multiple intrahepatic tumors with or without vascular spread, extrahepatic spread, or unstaged HCC. As has been done in previous studies of HCC outcomes using claims data,5 unstaged HCC was considered together with extrahepatic spread given the similar outcomes among these groups. These categories are similar to the tumor extent component of the Barcelona Clinic Liver Cancer staging system and approximate patients who may qualify for liver transplantation, surgical resection or locoregional therapies.19,20
County level economic and healthcare covariates were taken from the Area Health Resource File (AHRF)21 and the North Carolina Health Professional Data System. Measures were selected to cover domains used in other composite measures of census tract-based socioeconomic status (SES).22–24 Given the large number of covariates measured across 100 NC counties, we used factor analysis to create representative indices. Factor analysis was conducted separately for health system factors which together had an average variance extracted (AVE) of 58% across counties, and economic factors (see Supplemental Digital Content Table 1 for complete list and factor loadings). For economic variables, two factors combined for an AVE of 83.2% across counties. The first is described as economic disadvantage index as it was dominated by median home value, percent white, and unemployment rate. The second is described as the rurality index, as it was dominated by rurality and agricultural/ forestry/hunting/mining industries. For each index, the lowest quartile describes the least disadvantaged.
Additional patient level covariates were determined from ICD-9 claims in the 12 months pre-diagnosis. This included non-liver comorbidity using the Klaubunde modification of the Charlson Comorbidity Index (excluding liver disease),25 psychiatric comorbidity not including substance abuse (e.g. depression, anxiety, bipolar disorder, schizophrenia, post-traumatic stress disorder), liver-related complications (encephalopathy, ascites, varices, hepatorenal syndrome),26 and underlying cause of liver disease. Because many patients had low healthcare utilization in the year before diagnosis, codes for the cause of liver disease were evaluated from 12 months before to 1 month after diagnosis.
Pre-diagnosis healthcare utilization was determined by visits with a primary care provider or gastroenterology/hepatology provider (GI) in the year before diagnosis (excluding 2 months pre-diagnosis when consultations may reflect referrals for cancer), pre-diagnosis alpha fetoprotein (AFP) screening27 (guideline recommended every 6 months during this era for patients with cirrhosis and hepatitis virus),28 and number of unique contacts with the healthcare system (inpatient or outpatient visits). Consultation with HCC-specific subspecialties and at a liver transplant center was measured in the 90 days following diagnosis. Distance from each patient’s zip code to the closest liver transplant center was calculated. Treatment was defined as the initial treatment received as previously described,5 acknowledging that many patients go on to receive multiple therapies.
Analysis
We extracted patient and county level factors for all patients with HCC and linked claims data. We performed univariate analyses on demographic characteristics, insurance status, socioeconomic status, rurality, medical/psychiatric comorbidities, liver disease etiology and complications, and cancer stage at diagnosis, calculating median and interquartile range (IQR) for continuous variables and proportions for categorical variables. We calculated annual rates in subspecialty consultation within 90 days of HCC diagnosis and provision of HCC treatments. We performed multivariable logistic regression to assess variables associated with multifocal cancer at diagnosis, transplant center visit, and provision of HCC treatment. Lastly, we developed a Cox proportional hazards model to assess variables associated with overall survival. All statistical analyses were performed with SAS version 9.4 (SAS, Cary, NC, www.sas.com).
Results
Cohort Description
Our cohort included 1,809 patients with HCC (Table 1, Supplemental Digital Content Figure 1). The majority of patients had single (37%) or multiple (25%) tumors without vascular invasion or extrahepatic spread. A smaller proportion had vascular spread at diagnosis, including 6% and 13% of those with single and multiple lesions, respectively. The median age was 68 years (IQR 59–76) and patients were predominantly male (73%), white (76%), and insured by Medicare (59%).
Table 1:
Characteristics of patients with newly-diagnosed HCC
| HCC cases | |
|---|---|
| Characteristic | n=1,809 |
| Cancer Extent | |
| Single without vascular invasion | 671 (37%) |
| Single with vascular invasion | 105 (6%) |
| Multiple without vascular invasion | 458 (25%) |
| Multiple with vascular invasion | 230 (13%) |
| Extrahepatic | 66 (4%) |
| Not staged | 279 (15%) |
|
| |
| Age (median, interquartile range) | 68 (59, 76) |
| <50 | 110 (6%) |
| 50–64 | 574 (32%) |
| 65–74 | 611 (34%) |
| 75+ | 514 (28%) |
|
| |
| Sex | |
| Male | 1,326 (73%) |
| Female | 483 (27%) |
|
| |
| Race | |
| White | 1,375 (76%) |
| Black | 368 (20%) |
| Asian | Combined below† |
| Native American | Combined below† |
| Other | 66 (4%) |
|
| |
| Marital Status | |
| Married | 803 (44%) |
| Widowed | 33 (2%) |
| Divorced/separated/single | 640 (35%) |
| Other/unknown | 333 (18%) |
|
| |
| Insurance Payer at Diagnosis | |
| Private | 165 (9%) |
| Medicare | 1,069 (59%) |
| Medicaid/Dual | 575 (32%) |
| Tricare/VA | NA‡ |
| Other | NA‡ |
| Uninsured | NA‡ |
|
| |
| Psychiatric Comorbidity | |
| Yes | 820 (45%) |
| No | 989 (55%) |
|
| |
| Medical Comorbidity (CCI) | |
| 0 | 657 (36%) |
| 1 | 526 (29%) |
| 2+ | 626 (35%) |
|
| |
| Liver-related Complications | |
| 0 | 1,107 (61%) |
| 1+ | 702 (39%) |
|
| |
| Claims Restricted Cases | |
| Characteristic | n=1,809 |
|
| |
| Cause of Liver Disease§ | |
| Hepatitis B virus | 96 (5%) |
| Hepatitis C virus | 538 (30%) |
| Alcohol | 306 (17%) |
| Other cause of cirrhosis | 371 (21%) |
|
| |
| County Economic Disadvantage Index | |
| Quartile 1: Least Disadvantaged | 352 (19%) |
| Quartile 2 | 645 (36%) |
| Quartile 3 | 595 (33%) |
| Quartile 4: Most Disadvantaged | 216 (12%) |
|
| |
| County Rurality Index | |
| Quartile 1: Least Rural | 953 (53%) |
| Quartile 2 | 433 (24%) |
| Quartile 3 | 290 (16%) |
| Quartile 4: Most Rural | 132 (7%) |
|
| |
| County Healthcare Disadvantage Index | |
| Quartile 1: Least Disadvantaged | 897 (50%) |
| Quartile 2 | 389 (22%) |
| Quartile 3 | 308 (17%) |
| Quartile 4: Most Disadvantaged | 214 (12%) |
VA, Veterans Affairs; CCI, Charlson Comorbidity Index; PCP, primary care physician; GI, gastroenterology; AFP, alpha fetoprotein
To preserve confidentiality cells with <11 individuals were combined.
Claims not available for these insurers.
Causes of liver disease are not mutually exclusive, will not sum to 100%.
Multifocal HCC at Diagnosis
A total of 1,033 (57%) patients had multifocal HCC at presentation. Multifocal cancer at presentation was significantly associated with age, marital status, and sex. Receipt of pre-diagnosis AFP screening [adjusted odds ratio (aOR) of multifocal 0.72, 95% CI 0.58, 0.90] and GI care (aOR 0.77, 95% CI 0.61, 0.96) was associated with a decreased odds of multifocal disease (Table 2).
Table 2:
Association of stage at presentation and transplant center visit with patient and area-level factors
| All HCC Cases | Multifocal Cancer | Odds of Multifocal† | Transplant Center Visit | Odds of Transplant Center Visit† | |
|---|---|---|---|---|---|
| Characteristic | n=1,809 | n=1,033 | aOR, 95% CI | n = 957 | aOR, 95% CI |
| Cancer Extension | |||||
| Single without vascular invasion | 671 (37%) | NA | NA | 388 (41%) | Ref |
| Multiple without vascular invasion | 458 (25%) | NA | NA | 263 (27%) | 0.98 (0.74, 1.31) |
| Single with vascular invasion | 105 (6%) | NA | NA | 71 (7%) | 1.92 (1.16, 3.18) |
| Multiple with vascular invasion | 230 (13%) | NA | NA | 116 (12%) | 0.83 (0.58, 1.19) |
| Extrahepatic | 66 (4%) | NA | NA | 28 (3%) | 0.70 (0.39, 1.26) |
| Not staged | 279 (15%) | NA | NA | 91 (10%) | 0.34 (0.24, 0.49) |
|
| |||||
| Age (median) | 68 (59, 76) | 66 (57, 73) | |||
| <50 | 110 (6%) | 55 (5%) | Ref | 75 (8%) | Ref |
| 50–64 | 574 (32%) | 324 (31%) | 1.31 (0.85, 2.00) | 354 (37%) | 0.81 (0.48, 1.37) |
| 65–74 | 611 (34%) | 365 (35%) | 1.91 (1.19, 3.06) | 333 (35%) | 0.95 (0.54, 1.69) |
| 75+ | 514 (28%) | 289 (28%) | 1.64 (1.01, 2.68) | 195 (20%) | 0.52 (0.28, 0.94) |
|
| |||||
| Sex | |||||
| Male | 1,326 (73%) | 779 (75%) | Ref | 715 (75%) | Ref |
| Female | 483 (27%) | 254 (25%) | 0.74 (0.59, 0.93) | 242 (25%) | 0.95 (0.73, 1.23) |
|
| |||||
| Race | |||||
| White | 1,375 (76%) | 763 (74%) | Ref | 744 (78%) | Ref |
| Black | 368 (20%) | 232 (22%) | 1.22 (0.93, 1.61) | 174 (18%) | 0.54 (0.39, 0.74) |
| Other | 66 (4%) | 38 (4%) | 1.05 (0.62, 1.80) | 39 (4%) | 0.62 (0.32, 1.20) |
|
| |||||
| Marital Status | |||||
| Married | 803 (44%) | 429 (42%) | Ref | 427 (45%) | Ref |
| Widowed | 33 (2%) | 21 (2%) | 1.54 (0.73, 3.25) | 18 (2%) | 0.68 (0.29, 1.58) |
| Divorced/separated/single | 640 (35%) | 388 (38%) | 1.37 (1.08, 1.72) | 280 (29%) | 0.71 (0.54, 0.92) |
| Other/unknown | 333 (18%) | 195 (19%) | 1.26 (0.96, 1.65) | 232 (24%) | 2.12 (1.52, 2.98) |
|
| |||||
| Insurance Payer at Diagnosis | |||||
| Private | 165 (9%) | 87 (8%) | Ref | 129 (13%) | Ref |
| Medicare | 1,069 (59%) | 604 (58%) | 0.89 (0.60, 1.33) | 519 (54%) | 0.35 (0.21, 0.59) |
| Medicaid/Dual | 575 (32%) | 342 (33%) | 1.00 (0.68, 1.48) | 309 (32%) | 0.46 (0.28, 0.77) |
|
| |||||
| Medical Comorbidity (CCI) | |||||
| 0 | 657 (36%) | 389 (38%) | Ref | 377 (39%) | Ref |
| 1 | 526 (29%) | 299 (29%) | 0.95 (0.75, 1.22) | 283 (30%) | 0.80 (0.60, 1.07) |
| 2+ | 626 (35%) | 345 (33%) | 0.85 (0.67, 1.09) | 297 (31%) | 0.63 (0.48, 0.85) |
|
| |||||
| Cause of Liver Disease‡ | |||||
| Hepatitis B virus | 96 (5%) | 54 (5%) | 1.05 (0.67, 1.64) | 62 (6%) | 1.36 (0.80, 2.33) |
| Hepatitis C virus | 538 (30%) | 304 (29%) | 1.16 (0.89, 1.52) | 360 (38%) | 1.72 (1.25, 2.36) |
| Alcohol | 306 (17%) | 174 (17%) | 1.09 (0.81, 1.48) | 187 (20%) | 1.09 (0.76, 1.56) |
| Other cause of cirrhosis | 371 (21%) | 203 (20%) | 1.03 (0.78, 1.36) | 208 (22%) | 1.35 (0.98, 1.87) |
|
| |||||
| Prediagnosis Healthcare Utilization (median encounters, Q1, Q3) | |||||
| Tertile 1: 1 (0, 2) | 526 (29%) | 326 (32%) | Ref | 216 (23%) | Ref |
| Tertile 2: 5 (3, 6) | 657 (36%) | 377 (36%) | 0.95 (0.74, 1.21) | 352 (37%) | 1.45 (1.09, 1.94) |
| Tertile 3: 12 (9, 16) | 626 (35%) | 330 (32%) | 0.86 (0.66, 1.12) | 389 (41%) | 2.14 (1.57, 2.93) |
|
| |||||
| 1 year prediagnosis specialty care | |||||
| None | 245 (14%) | 725 (70%) | Ref | 419 (44%) | Ref |
| PCP only | 951 (53%) | § | § | § | § |
| PCP + GI/hepatology | 567 (31%) | 308 (30%) | 0.77 (0.61, 0.96) | 538 (56%) | 1.66 (1.27, 2.18) |
| GI/Hepatology only | 46 (3%) | § | § | § | § |
|
| |||||
| 1 year prediagnosis AFP Screening | |||||
| Yes | 779 (43%) | 398 (39%) | 0.72 (0.58, 0.90) | 521 (54%) | 1.74 (1.35, 2.23) |
| No | 1,030 (57%) | 635 (61%) | Ref | 436 (46%) | Ref |
|
| |||||
| Miles to Closest Liver Txp Center (median, Q1, Q3) | NA | ||||
| Tertile 1: Median=17.4 (7.3, 25.3) | 605 (33%) | 346 (34%) | 442 (46%) | Ref | |
| Tertile 2: Median=53.(44.2, 60.8) | 602 (33%) | 354 (34%) | 268 (28%) | 0.46 (0.32, 0.66) | |
| Tertile 3: Median=106 (90., 133) | 600 (33%) | 331 (32%) | 247 (26%) | 0.27 (0.18, 0.41) | |
|
| |||||
| County Economic Disadvantage Index | |||||
| Quartile 1: Least Disadvantaged | 352 (19%) | 185 (18%) | Ref | 164 (17%) | Ref |
| Quartile 2 | 645 (36%) | 369 (36%) | 1.13 (0.85, 1.51) | 427 (45%) | 0.96 (0.65, 1.41) |
| Quartile 3 | 595 (33%) | 337 (33%) | 1.06 (0.77, 1.45) | 248 (26%) | 0.36 (0.24, 0.56) |
| Quartile 4: Most Disadvantaged | 216 (12%) | 141 (14%) | 1.50 (1.01, 2.23) | 118 (12%) | 1.76 (1.10, 2.83) |
|
| |||||
| County Rurality Index | |||||
| Quartile 1: Least Rural | 953 (53%) | 550 (53%) | Ref | 553 (58%) | Ref |
| Quartile 2 | 433 (24%) | 240 (23%) | 0.87 (0.65, 1.77) | 214 (22%) | 0.56 (0.39, 0.81) |
| Quartile 3 | 290 (16%) | 165 (16%) | 0.89 (0.65, 1.20) | 158 (17%) | 1.28 (0.84, 1.96) |
| Quartile 4: Most Rural | 132 (7%) | 77 (7%) | 0.90 (0.58, 1.39) | 32 (3%) | 0.24 (0.14, 0.47) |
|
| |||||
| County Healthcare Disadvantage Index | |||||
| Quartile 1: Least Disadvantaged | 853 (47%) | 478 (46%) | Ref | 469 (49%) | Ref |
| Quartile 2 | 441 (24%) | 258 (25%) | 1.06 (0.81, 1.39) | 241 (25%) | 0.98 (0.70, 1.37) |
| Quartile 3 | 296 (16%) | 164 (16%) | 1.04 (0.78, 1.38) | 151 (16%) | 1.01 (070, 1.45) |
| Quartile 4: Most Disadvantaged | 218 (12%) | 132 (13%) | 1.21 (0.86, 1.71) | 96 (10%) | 0.73 (0.48, 1.10) |
CCI, Charlson Comorbidity Index; PCP, primary care physician; GI, gastroenterology; AFP, alpha fetoprotein
Model was also adjusted for psychiatric comorbidity, liver comorbidity (complications of cirrhosis), number of hepatologists per 100,000 population, which were not significantly associated with survival with minimal/no trend suggesting possible effect. These variables were omitted to condense the table size.
Causes of liver disease are not mutually exclusive, will not sum to 100%.
Cells combined with cell above for multivariable model, comparing “any GI/hepatology” to referent of “none or PCP only”.
Visit at a Liver Transplant Center
In the 90 days following diagnosis, 957 (53%) of patients were seen at a liver transplant center for any reason (Table 2, Figure 1). A transplant center visit was significantly less likely among older patients (aOR ≥75 versus <50 years 0.52, 95% CI 0.28, 0.94), black patients (aOR versus whites 0.54, 95% CI 0.39, 0.74), and patients with Medicare (aOR versus private 0.35, 95% CI 0.21, 0.59) and Medicaid (aOR versus private 0.46, 95% CI 0.28, 0.77).
Figure 1: Time trends of subspecialty consultation within 90 days of diagnosis among incident HCC cases diagnosed 2004–2012.
The proportion of all HCC cases diagnosed from 2004–2012 with a visit at a transplant center or with a consultation by a surgeon, hematologist/oncologist (HEM/ONC), gastroenterology/hepatology (GI/HEPATOLOGY) within 90 days of diagnosis.
Patients residing a median of 53 miles (aOR 0.46, 95% CI 0.32, 0.66) and a median of 106 miles (aOR 0.27, 95% CI 0.18, 0.41) were less likely to be seen at a transplant center than those living in the closest tertile (median 17 miles). Transplant center visits were also less likely among those living in the most rural counties (aOR versus least rural 0.24, 95% CI 0.14, 0.47). Pre-diagnosis healthcare was also a major determinant of transplant center visit, with an increased odds among those with more pre-diagnosis healthcare utilization (aOR highest versus lowest tertile 2.14, 95% CI 1.57, 2.93), AFP testing (aOR 1.74, 95% CI 1.35, 2.23), and GI care (aOR 1.66, 95% CI 1.27, 2.18).
Provision of HCC Treatment
Of 1,809 HCC patients, 30% died within the 90 day treatment exposure window following diagnosis and were excluded from treatment and survival analyses. These patients were older with greater comorbidity, more advanced cancer, more likely to be divorced, and were less likely to have received pre-diagnosis care. They were significantly less likely to be seen at a transplant center or receive treatment for their HCC.
In the 1,250 patients surviving the 90 day treatment exposure window, 857 (69%) were treated (Table 3). Of these, 478 (56%) has a surgical consultation, 425 (50%) consultation with GI, and 469 (55%) saw a hematologist/oncologist and these rates slowly increased from 2003 to 2013 (Figure 1).
Table 3:
Factors associated with treatment among HCC patients surviving 90 days from diagnosis
| Treated N=857 (69%) | Untreated N=393 (31%) | Odds of Treatment† | ||
|---|---|---|---|---|
| Characteristic | n (%) | n (%) | OR, 95% CI | aOR, 95% CI |
| Initial Treatment | ||||
| Curative surgery | 149 (12%) | - | - | - |
| Ablation | 186 (15%) | - | - | - |
| LRT (TACE + TARE) | 387 (31%) | - | - | - |
| Drug Therapy | 89 (7%) | - | - | - |
| Radiation | 46 (4%) | - | - | - |
|
| ||||
| Year of Diagnosis | ||||
| 2004–2005 | 162 (19%) | 78 (20%) | Ref | Ref |
| 2006–2007 | 160 (19%) | 75 (19%) | 1.03 (0.70, 1.51) | 0.91 (0.58, 1.44) |
| 2008–2009 | 200 (23%) | 101 (26%) | 0.95 (0.66, 1.37) | 0.86 (0.56, 1.33) |
| 2010–2011 | 208 (24%) | 77 (20%) | 1.30 (0.89, 1.89) | 1.33 (0.85, 2.07) |
| 2012–2013 | 127 (15%) | 62 (16%) | 0.99 (0.66, 1.48) | 1.09 (0.65, 1.82) |
|
| ||||
| Age (median) | ||||
| <65 | 382 (45%) | 134 (34%) | Ref | Ref |
| 65–74 | 306 (36%) | 112 (28%) | 0.96 (0.72, 1.28) | 0.64 (0.41, 0.99) |
| 75+ | 169 (20%) | 147 (37%) | 0.40 (0.30, 0.54) | 0.30 (0.18, 0.49) |
|
| ||||
| Race | ||||
| White | 661 (77%) | 289 (74%) | Ref | Ref |
| Black | 162 (19%) | 92 (23%) | 0.77 (0.58, 1.03) | 0.84 (0.56, 1.24) |
| Other | 34 (4%) | 12 (3%) | 1.24 (0.63, 2.43) | 1.42 (0.64, 3.18) |
|
| ||||
| Marital Status | ||||
| Married | 415 (48%) | 150 (38%) | Ref | Ref |
| Divorced/separated/ Single | 239 (28%) | 173 (44%) | 0.53 (0.41, 0.69) | 0.62 (0.44, 0.88) |
| Other/unknown | 203 (24%) | 70 (18%) | 0.99 (0.71, 1.38) | 0.81 (0.53, 1.22) |
|
| ||||
| Insurance Payer at Diagnosis | ||||
| Private | 117 (14%) | 18 (5%) | Ref | Ref |
| Medicare | 478 (56%) | 233 (59%) | 0.32 (0.19, 0.53) | 0.78 (0.40, 1.50) |
| Medicaid/Dual | 262 (31%) | 142 (36%) | 0.28 (0.17, 0.49) | 0.80 (0.41, 1.54) |
|
| ||||
| Psychiatric comorbidity | ||||
| No | 488 (57%) | 192 (49%) | Ref | Ref |
| Yes | 369 (43%) | 201 (51%) | 0.72 (0.57, 0.92) | 0.64 (0.47, 0.89) |
|
| ||||
| Medical Comorbidity (CCI) | ||||
| 0 | 354 (41%) | 140 (36%) | Ref | Ref |
| 1 | 266 (31%) | 108 (27%) | 0.97 (0.72, 1.31) | 0.90 (0.63, 1.30) |
| 2+ | 237 (28%) | 145 (37%) | 0.65 (0.49, 0.86) | 0.78 (0.54, 1.12) |
|
| ||||
| Treated | Untreated | Odds of Treatment† | ||
| Characteristic | n (%) | n (%) | OR, 95% CI | aOR, 95% CI |
|
| ||||
| Liver-related Complications | ||||
| 0 | 561 (65%) | 250 (64%) | Ref | Ref |
| 1+ | 296 (35%) | 143 (36%) | 0.92 (0.72, 1.18) | 0.51 (0.35, 0.74) |
|
| ||||
| Cancer Extension | ||||
| Single without vascular invasion | 375 (44%) | 129 (33%) | Ref | Ref |
| Single with vascular invasion | 59 (7%) | 28 (7%) | 0.72 (0.44, 1.19) | 0.69 (0.38, 1.24) |
| Multiple without vascular invasion | 249 (29%) | 87 (22%) | 0.98 (0.72, 1.35) | 1.07 (0.74, 1.54) |
| Multiple with Vascular invasion | 88 (10%) | 52 (13%) | 0.58 (0.39, 0.87) | 0.59 (0.37, 0.94) |
| Extrahepatic | 20 (2%) | 15 (4%) | 0.46 (0.23, 0.92) | 0.43 (0.19, 0.99) |
| Not staged | 66 (8%) | 82 (21%) | 0.28 (0.19, 0.41) | 0.43 (0.28, 0.68) |
|
| ||||
| 1 year Prediagnosis Specialty Care | ||||
| None | 86 (10%) | 65 (17%) | Ref | Ref |
| PCP only | 412 (48%) | 225 (57%) | 1.38 (0.96, 1.99) | 1.36 (0.85, 2.17) |
| GI/Hepatology | 359 (42%) | 103 (26%) | 2.63 (1.78, 3.89) | 1.68 (0.98, 2.86) |
|
| ||||
| 1 year Prediagnosis AFP Screening | ||||
| No | 354 (41%) | 267 (68%) | Ref | Ref |
| Yes | 503 (59%) | 126 (32%) | 3.01 (2.34, 3.87) | 2.61 (1.90, 3.60) |
|
| ||||
| Surgical Consult in 90 days after diagnosis | ||||
| No | 379 (44%) | 295 (75%) | Ref | Ref |
| Yes | 478 (56%) | 98 (25%) | 3.80 (2.91, 4.95) | 3.40 (2.48, 4.67) |
|
| ||||
| Visit at Liver Transplant Center in 90 days from Diagnosis | ||||
| Yes | 659 (77%) | 183 (47%) | 3.54 (2.76, 4.54) | 2.42 (1.74, 3.36) |
| No | 198 (23%) | 210 (53%) | Ref | Ref |
|
| ||||
| County Economic Disadvantage Index | ||||
| Quartile 1: Least Disadvantaged | 175 (20%) | 66 (17%) | Ref | Ref |
| Quartile 2 | 307 (36%) | 143 (36%) | 0.81 (0.57, 1.14) | 0.56 (0.35, 0.87) |
| Quartile 3 | 277 (32%) | 135 (34%) | 0.77 (0.55, 1.10) | 0.80 (0.50, 1.30) |
| Quartile 4: Most Disadvantaged | 98 (11%) | 49 (12%) | 0.75 (0.48, 1.18) | 0.79 (0.44, 1.43) |
|
| ||||
| County Rurality Index | ||||
| Quartile 1: Least Rural | 467 (54%) | 203 (52%) | Ref | Ref |
| Quartile 2 | 197 (23%) | 92 (23%) | 0.93 (0.69, 1.25) | 0.90 (0.58, 1.39) |
| Quartile 3 | 139 (16%) | 64 (16%) | 0.94 (0.67, 1.32) | 0.70 (0.45, 1.11) |
| Quartile 4: Most Rural | 54 (6%) | 34 (9%) | 0.69 (0.44, 1.09) | 0.70 (0.37, 1.33) |
CCI, Charlson Comorbidity Index; PCP, primary care physician; GI, gastroenterology; AFP, alpha fetoprotein
Model was also adjusted for sex, cause of liver disease, prediagnosis healthcare utilization, post-diagnosis GI and Hematology/Oncology consultation, county health services index, which were not significantly associated with survival with minimal/no trend suggesting possible effect. These variables were omitted to condense the table size.
Factors most strongly associated with receipt of HCC treatment included pre-diagnosis AFP screening (aOR 2.61, 95% CI 1.90, 3.60), surgical consultation (aOR 3.40, 95% CI 2.48, 4.67) and visit at a liver transplant center (aOR 2.42, 95% CI 1.74, 3.36) (Table 4). In contrast, advanced age, unmarried status, psychiatric comorbidity, and complications of cirrhosis all significantly reduced the odds of receiving HCC treatment.
Table 4:
Factors associated with survival among newly diagnosed cases of HCC
| Cases surviving 90 days n=1,250 | Hazard for Mortality† | ||
|---|---|---|---|
| Patient Level Characteristic | n (%) | HR,95% CI | aHR,95% CI |
| Year | |||
| 2004–2007 | 475 (38%) | Ref | Ref |
| 2008–2013 | 775 (62%) | 0.85 (0.75, 0.97) | 0.75 (0.66, 0.85) |
|
| |||
| Age | |||
| <50 | 93 (7%) | Ref | Ref |
| 50–64 | 423 (34%) | 0.97 (0.76, 1.26) | 1.08 (0.83, 1.40) |
| 65–74 | 418 (33%) | 1.32 (1.02, 1.69) | 1.34 (1.00, 1.78) |
| 75+ | 316 (25%) | 1.48 (1.15, 1.92) | 1.39 (1.03, 1.87) |
|
| |||
| Race | |||
| White | 950 (76%) | Ref | Ref |
| Black | 254 (20%) | 1.09 (0.94, 1.27) | 1.09 (0.91, 1.30) |
| Other | 46 (4%) | 0.92 (0.66, 1.28) | 1.40 (0.98, 1.99) |
|
| |||
| Insurance Payer at Dx | |||
| Private | 135 (11%) | Ref | Ref |
| Medicare | 711 (57%) | 2.31 (1.82, 2.93) | 1.58 (1.20, 2.09) |
| Medicaid/Dual | 404 (32%) | 2.50 (1.95, 3.20) | 1.55 (1.17, 2.05) |
|
| |||
| Psychiatric Comorbidity | |||
| No | 680 (54%) | Ref | Ref |
| Yes | 570 (46%) | 1.16 (1.03, 1.31) | 1.15 (1.00, 1.32) |
|
| |||
| Medical Comorbidity (CCI) | |||
| 0 | 494 (40%) | Ref | Ref |
| 1 | 374 (30%) | 1.18 (1.02, 1.36) | 1.14 (0.97, 1.33) |
| 2+ | 382 (31%) | 1.40 (1.21, 1.62) | 1.18 (1.00, 1.38) |
|
| |||
| Liver-related Complications | |||
| 0 | 811 (65%) | Ref | Ref |
| 1+ | 439 (35%) | 1.00 (0.88, 1.14) | 1.24 (1.05, 1.45) |
|
| |||
| Cancer Extension | |||
| Single lesion | 591 (47%) | Ref | Ref |
| Multiple without vascular invasion | 336 (27%) | 1.51 (1.31, 1.75) | 1.41 (1.21, 1.64) |
| Multiple with Vascular invasion | 140 (11%) | 2.18 (1.79, 2.66) | 1.74 (1.42, 2.14) |
| Extrahepatic/not staged | 183 (15%) | 2.05 (1.71, 2.44) | 1.39 (1.15, 1.69) |
|
| |||
| Initial Cancer Directed Treatment | |||
| Curative Surgery | 149 (12%) | Ref | Ref |
| Ablation | 186 (15%) | 1.50 (1.14, 1.97) | 1.60 (1.20, 2.13) |
| LRT (TACE/Y90) | 387 (31%) | 2.68 (2.11, 3.41) | 2.47 (1.90, 3.19) |
| Drug Therapy (including Sorafenib) | 89 (7%) | 5.81 (4.28, 7.90) | 4.57 (3.27, 6.40) |
| Radiation | 46 (4%) | 4.49 (3.10, 6.49) | 3.79 (2.55, 5.65) |
| Never Treated | 393 (31%) | 5.57 (4.38, 7.09) | 4.97 (3.79, 6.51) |
|
| |||
| 1 year Prediagnosis Specialty Care | |||
| None | 151 (12%) | Ref | Ref |
| PCP only | 637 (51%) | 0.94 (0.78, 1.13) | 0.83 (0.67, 1.03) |
| GI/Hepatology | 462 (37%) | 0.68 (0.56–0.83) | 0.74 (0.58, 0.94) |
|
| |||
| Hematology/Oncology Consult | |||
| No | 562 (45%) | Ref | Ref |
| Yes | 688 (55%) | 1.53 (1.36, 1.73) | 1.39 (1.21, 1.59) |
CCI, Charlson Comorbidity Index; PCP, primary care physician; GI, gastroenterology.
Model was also adjusted for sex, marital status, cause of liver disease, prediagnosis healthcare utilization, prediagnosis AFP screening, surgical or GI/hepatology consult in 90 days of diagnosis, National Cancer Institute Center or liver transplant center visit in 90 days after diagnosis, and county level economic, rurality, and health services disadvantage indices, which were not significantly associated with survival with minimal/no trend suggesting possible effect. These variables were omitted to condense the table size.
Overall Survival
Across the study period, survival increased from a median survival of 6 months (interquartile range (IQR) 2, 21) in 2004–2006 to 8 months (IQR 2, 28) in 2010–2012. In adjusted models accounting for disease severity and treatment, patients diagnosed in 2008 and beyond had significantly better survival than patients diagnosed 2004–2007 [adjusted hazard ratio (aHR) for death 0.75, 95% CI 0.66, 0.85].
In addition to year of diagnosis, cancer stage and receipt of cancer treatment were strongly associated with survival: compared with the 12% of patients who underwent curative surgery, the risk of death was higher among patients treated with ablation (aHR 1.60, 95% CI 1.20, 2.13) and LRT (aHR 2.47, 95% CI 1.90, 3.19). Patients treated with drug therapy (aHR 4.57, 95% CI 3.27, 6.40), radiation (aHR 3.79, 95% CI 2.55, 5.65), and untreated patients (aHR 4.97, 95% CI 3.79, 6.51) all had a higher risk of mortality, although these were not adjusted for factors that may influence treatment selection (e.g. bilirubin) (Table 4).
Survival was better among younger patients and those without comorbidities. Receipt of pre-diagnosis GI care was associated with improved survival compared to patients who saw neither PCP nor GI in the year before diagnosis, (aHR 0.74, 95% CI 0.58, 0.94). Despite adjusting for age and treatment received, patients with Medicare (aHR 1.58, 95% CI 1.20, 2.09) and Medicaid (aHR 1.55, 95% CI 1.17, 2.05) had significantly worse survival than privately insured patients.
Discussion
In this population-based retrospective cohort study examining the effects of patient characteristics, county level resources, and healthcare utilization on HCC outcomes, we found patient-level sociodemographic factors (older age, black race, unmarried status, insurance) to be key determinants of stage at diagnosis and survival following an HCC diagnosis. These same patient factors were associated with liver transplant center visits and cancer-directed treatment. When further analyzing the root causes, survival was most strongly associated with receipt of cancer-directed treatment. Treatment, in turn, was more likely in patients with pre-diagnosis specialty GI care and screening, and post-diagnosis liver transplant center visits, which are likely a surrogate for receipt of multidisciplinary care. The disparities in outcomes are therefore largely accounted for by lower quality healthcare before and after an HCC diagnosis. In all analyses, Medicaid and Medicare beneficiaries experienced significantly inferior outcomes, including a marked reduction in survival, compared to privately insured HCC patients.
Given that HCC incidence is higher among racial and ethnic minorities and in socioeconomically disadvantaged regions,29,30 we hypothesized that a lack of access to local healthcare resources and greater SES disadvantage are key contributing factors to the low rates of treatment and disparities in care which have been previously reported. After investigating a wide array of county level determinants of economic health and healthcare availability, the only clear associations between county factors and the quality of HCC care was that patients residing in the most rural counties and counties farthest from a transplant center were least likely to be seen at a transplant center in the 90 days after diagnosis. No other county-level factors clearly influenced treatment or survival in HCC.
Health insurance was strongly associated with the likelihood of transplant center visit and survival. The association between Medicare insurance and these outcomes could be confounded by age, given that Medicare beneficiaries are usually >65 years old and some patients may be ineligible for liver transplantation based on advanced age. However, recent data suggests an increased proportion of patients older than 65 years with HCC are being listed for transplantation31 and the negative associations between Medicare insurance and transplant center visit and survival persisted in multivariable models adjusting for age. Patients with Medicare and Medicaid had a 24% and 57% increased risk of death compared to the privately insured; it was 67% higher for the uninsured. The effect of insurance status on survival did not diminish after adjusting for patient age, comorbidity and treatment with a 58% and 55% increase in the risk of death for Medicare and Medicaid patients. The influence of public insurance on outcomes could reflect limitations imposed by Medicare and Medicaid on access to care, however, public insurance in this analysis is also likely a proxy for individual socioeconomic resources given that income limits for NC Medicaid are strict.32 Regardless of county resources, such individuals likely face additional financial challenges that could compromise access compared with their privately insured counterparts. The marked and persistently inferior outcome among single, divorced, or widowed patients (advanced stage at presentation, less consultative care and treatment) and those with psychiatric comorbidity (lower rates of treatment and inferior survival) also speak to the importance of individual resources (e.g. social support) on the ability of HCC patients to receive the care they need.
This study is strengthened by its large population-based sample and the availability of linked insurance claims, allowing for the simultaneous exploration of patient, treatment and facility factors on outcomes. However, this study must be interpreted in the context of potential limitations. First, the survival analysis is limited to patients who survived the 90 day treatment exposure window following diagnosis. While this restriction was necessary to evaluate the effect of treatment on survival, omitting these sickest patients may have obscured the effect of county economic and healthcare factors on HCC outcomes. Though we generally found that the same factors that were associated with treatment were associated with early mortality (advanced age and stage, greater comorbidity, single/divorced, Medicare and Medicaid, less pre-diagnostic care), one potentially meaningful difference was that patients with early mortality were more likely to live in a county with fewer healthcare services and lower density of GI physicians. Second, the data source did not allow for adjustment for liver disease severity via the Model for End Stage Liver Disease (MELD) or Child-Pugh score, although we did account for pre-diagnosis liver-related complications. Lastly, our data source did not allow us to assess for factors that could have contributed to improved survival over time including improved treatments for underlying liver disease (e.g. direct acting antiviral therapy for hepatitis C virus), expanded access to HCC therapies and clinical trials, and broader adoption of multidisciplinary tumor boards and clinics.
Our findings that an individual’s use of the healthcare system before diagnosis and visits at an expert center early in their cancer course are major driving forces behind treatment and survival for HCC are critical when thinking of how to improve outcomes of HCC patients—in our state and across the nation. To reduce the rates of very early mortality from HCC, public health efforts must focus on detection of cirrhosis and GI referrals for affected individuals. To improve survival among patients who present earlier in their disease course, efforts must focus on increasing access to subspecialty care and treatment. Our ongoing work will examine patient reported barriers to accessing care following an HCC diagnosis, with a focus on high risk black and rural residents, and Medicare and Medicaid beneficiaries.
Supplementary Material
Figure 2: Time trends of type of initial treatment received by year of diagnosis among incident HCC cases diagnosed 2004–2012.
The type of first treatment received among patients with HCC diagnosed from 2004–2012. The majority of patients diagnosed with HCC received no treatment for every year during the study period; TACE: transarterial chemoembolization; Y90: yttrium-90
Grant Support:
This work was supported by National Cancer Institute, K07CA160722 (HKS) and National Institutes of Health grant T32 DK007634 (AMM). Additional support was provided by the Cancer information & Population Health Resource (CIPHR), UNC Lineberger Comprehensive Cancer Center with funding provided by the University Cancer Research Fund (UCRF) via the State of North Carolina
Abbreviations:
- AFP
alpha fetoprotein
- aHR
adjusted hazard ratio
- aOR
adjusted odds ratio
- AVE
average variance extracted
- CCI
Charlson Comorbidity Index
- GI
gastroenterology
- HRF
Area Health Resource File
- HCC
hepatocellular carcinoma
- ICD
International Classification of Diseases
- IQR
interquartile range
- LRTs
locoregional therapies
- NCCCR
North Carolina Central Cancer Registry
- PCP
primary care physician
- SES
socioeconomic status
- TACE
transarterial chemoemoblization
- VA
Veterans Affairs
- Y90
yttrium-90
Footnotes
Conflicts of Interest/Disclosures: Dr. Sanoff has received research funding paid to the University of North Carolina from Bayer. Dr. Lund’s spouse is an employee of GlaxoSmithKline. Dr. Barritt has received research funding paid to the University of North Carolina from Intercept Pharmaceuticals, Genfit Pharmaceuticals, Bristol Myers Squibb, NuSirt, Target Pharamsolutions. Drs. Moon, Chang, Hayashi, and Stitzenberg report no conflicts.
References
- 1.White DL, Thrift AP, Kanwal F, Davila J, El-Serag HB. Incidence of Hepatocellular Carcinoma in All 50 United States, From 2000 Through 2012. Gastroenterology. 2017;152(4):812–820 e815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Petrick JL, Kelly SP, Altekruse SF, McGlynn KA, Rosenberg PS. Future of Hepatocellular Carcinoma Incidence in the United States Forecast Through 2030. J Clin Oncol. 2016;34(15):1787–1794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kohler BA, Sherman RL, Howlader N, et al. Annual Report to the Nation on the Status of Cancer, 1975–2011, Featuring Incidence of Breast Cancer Subtypes by Race/Ethnicity, Poverty, and State. Journal of the National Cancer Institute. 2015;107(6):djv048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.El-Serag HB, Siegel AB, Davila JA, et al. Treatment and outcomes of treating of hepatocellular carcinoma among Medicare recipients in the United States: a population-based study. J Hepatol. 2006;44(1):158–166. [DOI] [PubMed] [Google Scholar]
- 5.Sanoff HK, Chang Y, Stavas JM, Sturmer T, Lund J. Effectiveness of Initial Transarterial Chemoembolization for Hepatocellular Carcinoma Among Medicare Beneficiaries. J Natl Compr Canc Netw. 2015;13(9):1102–1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Davila JA, Kramer JR, Duan Z, et al. Referral and receipt of treatment for hepatocellular carcinoma in United States veterans: effect of patient and nonpatient factors. Hepatology. 2013;57(5):1858–1868. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shah SA, Smith JK, Li Y, Ng SC, Carroll JE, Tseng JF. Underutilization of therapy for hepatocellular carcinoma in the medicare population. Cancer. 2011;117(5):1019–1026. [DOI] [PubMed] [Google Scholar]
- 8.Lopez PM, Villanueva A, Llovet JM. Systematic review: evidence-based management of hepatocellular carcinoma--an updated analysis of randomized controlled trials. Aliment Pharmacol Ther. 2006;23(11):1535–1547. [DOI] [PubMed] [Google Scholar]
- 9.Wang J, Ha J, Lopez A, Bhuket T, Liu B, Wong RJ. Medicaid and Uninsured Hepatocellular Carcinoma Patients Have More Advanced Tumor Stage and Are Less Likely to Receive Treatment. J Clin Gastroenterol. 2018;52(5):437–443. [DOI] [PubMed] [Google Scholar]
- 10.Mokdad AA, Zhu H, Marrero JA, Mansour JC, Singal AG, Yopp AC. Hospital Volume and Survival After Hepatocellular Carcinoma Diagnosis. Am J Gastroenterol. 2016;111(7):967–975. [DOI] [PubMed] [Google Scholar]
- 11.Tan D, Yopp A, Beg MS, Gopal P, Singal AG. Meta-analysis: underutilisation and disparities of treatment among patients with hepatocellular carcinoma in the United States. Aliment Pharmacol Ther. 2013;38(7):703–712. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Artinyan A, Mailey B, Sanchez-Luege N, et al. Race, ethnicity, and socioeconomic status influence the survival of patients with hepatocellular carcinoma in the United States. Cancer. 2010;116(5):1367–1377. [DOI] [PubMed] [Google Scholar]
- 13.Davila JA, El-Serag HB. Racial differences in survival of hepatocellular carcinoma in the United States: a population-based study. Clin Gastroenterol Hepatol. 2006;4(1):104–110; quiz 104–105. [PubMed] [Google Scholar]
- 14.Siegel AB, McBride RB, El-Serag HB, et al. Racial disparities in utilization of liver transplantation for hepatocellular carcinoma in the United States, 1998–2002. The American journal of gastroenterology. 2008;103(1):120–127. [DOI] [PubMed] [Google Scholar]
- 15.Davila JA, Duan Z, McGlynn KA, El-Serag HB. Utilization and outcomes of palliative therapy for hepatocellular carcinoma: a population-based study in the United States. J Clin Gastroenterol. 2012;46(1):71–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rich NE, Hester C, Odewole M, et al. Racial and Ethnic Differences in Presentation and Outcomes of Hepatocellular Carcinoma. Clin Gastroenterol Hepatol. 2019;17(3):551–559 e551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Serper M, Taddei TH, Mehta R, et al. Association of Provider Specialty and Multidisciplinary Care With Hepatocellular Carcinoma Treatment and Mortality. Gastroenterology. 2017;152(8):1954–1964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Meyer AM, Olshan AF, Green L, et al. Big data for population-based cancer research: the integrated cancer information and surveillance system. North Carolina medical journal. 2014;75(4):265–269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis. 2010;30(1):61–74. [DOI] [PubMed] [Google Scholar]
- 20.Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis. 1999;19(3):329–338. [DOI] [PubMed] [Google Scholar]
- 21.HRSA HRaSA. Area Health Resources File. National, State and County Health Resources Information Database. http://ahrf.hrsa.gov/index.htm. Published 2016. Accessed 9 November, 2016.
- 22.Yu M, Tatalovich Z, Gibson JT, Cronin KA. Using a composite index of socioeconomic status to investigate health disparities while protecting the confidentiality of cancer registry data. Cancer causes & control : CCC. 2014;25(1):81–92. [DOI] [PubMed] [Google Scholar]
- 23.Yost K, Perkins C, Cohen R, Morris C, Wright W. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer causes & control : CCC. 2001;12(8):703–711. [DOI] [PubMed] [Google Scholar]
- 24.Krieger N, Chen JT, Waterman PD, Soobader MJ, Subramanian SV, Carson R. Geocoding and monitoring of US socioeconomic inequalities in mortality and cancer incidence: does the choice of area-based measure and geographic level matter?: the Public Health Disparities Geocoding Project. American journal of epidemiology. 2002;156(5):471–482. [DOI] [PubMed] [Google Scholar]
- 25.Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. Journal of clinical epidemiology. 2000;53(12):1258–1267. [DOI] [PubMed] [Google Scholar]
- 26.Ulahannan SV, Duffy AG, McNeel TS, et al. Earlier presentation and application of curative treatments in hepatocellular carcinoma. Hepatology. 2014;60(5):1637–1644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Davila JA, Morgan RO, Richardson PA, Du XL, McGlynn KA, El-Serag HB. Use of surveillance for hepatocellular carcinoma among patients with cirrhosis in the United States. Hepatology. 2010;52(1):132–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bruix J, Sherman M. AASLD Practice Guideline Management of Hepatocellular Carcinoma: An Update. Hepatology. 2010;published online only; July 2010. www.aasld.org. [Google Scholar]
- 29.Shebl FM, Capo-Ramos DE, Graubard BI, McGlynn KA, Altekruse SF. Socioeconomic status and hepatocellular carcinoma in the United States. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2012;21(8):1330–1335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chang ET, Yang J, Alfaro-Velcamp T, So SK, Glaser SL, Gomez SL. Disparities in liver cancer incidence by nativity, acculturation, and socioeconomic status in California Hispanics and Asians. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2010;19(12):3106–3118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cullaro G, Rubin JB, Mehta N, Lai JC. Differential Impact of Age Among Liver Transplant Candidates With and Without Hepatocellular Carcinoma. Liver Transpl. 2020;26(3):349–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.NC Department of Health and Human Services, Medical Assistance. “Eligibility for Medicaid or Health Choice”. http://dma.ncdhhs.gov/medicaid/get-started/eligibility-for-medicaid-or-health-choice Accessed 18 Jan, 2017. Accessed.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.