24.12.2012 Views

Annual Meeting Proceedings Part 1 - American Society of Clinical ...

Annual Meeting Proceedings Part 1 - American Society of Clinical ...

Annual Meeting Proceedings Part 1 - American Society of Clinical ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

412s Health Services Research<br />

6122 General Poster Session (Board #11G), Mon, 1:15 PM-5:15 PM<br />

Variation in health-related quality <strong>of</strong> life (HRQOL) by ECOG performance<br />

status (PS) and fatigue among patients with chronic lymphocytic leukemia<br />

(CLL). Presenting Author: Christopher Flowers, Emory University, Atlanta,<br />

GA<br />

Background: Clinicians and investigators commonly use ECOG PS and<br />

clinician-reported patient (pt) fatigue as surrogates for HRQOL, a multifaceted<br />

construct that comprehensively looks at the pt perspective on<br />

disease and well-being. Because limited data exist on the relationships<br />

between PS, fatigue, and HRQOL for CLL pts, we examined the associations<br />

between these measures, and 3 validated HRQOL instruments: the<br />

Functional Assessment <strong>of</strong> Cancer Therapy-Leukemia (FACT-Leu), EQ-5D,<br />

and Brief Fatigue Inventory (BFI). Methods: Data were collected in<br />

CONNECT CLL, a prospective US observational registry initiated in 2010.<br />

Patient demographics and clinical characteristics were provided by clinicians.<br />

Patient HRQOL was self-reported at enrollment using the FACT-Leu,<br />

EQ-5D, and BFI. Scores were analyzed by ECOG PS (0, 1, 2-4) and<br />

clinician-reported fatigue (yes, no). Differences in HRQOL scores were<br />

assessed by ANOVA. Results: HRQOL data were reported by 899 pts from<br />

148 community, 10 academic, and 3 government centers. ECOG PS was<br />

available on 711 pts. Overall HRQOL, measured by mean FACT-Leu,<br />

FACT-G and EQ-5D Visual Analogue Scale (VAS), worsened with ECOG PS<br />

severity and was worse in pts with fatigue (all p�0.0001). All FACT-Leu<br />

domains except social/family were worse in pts with fatigue and those with<br />

higher ECOG PS. Mean EQ-5D pain/discomfort, mobility, self care and<br />

usual activities domain scores worsened in severity as ECOG worsened and<br />

for pts with fatigue (all p�0.011). BFI data indicated that global fatigue,<br />

fatigue severity and fatigue-related interference worsened by ECOG severity<br />

and were associated with clinician-reported fatigue (all p�0.0001).<br />

Conclusions: Initial CONNECT CLL results confirm that HRQOL worsens<br />

with worsening ECOG PS and was worse among pts with fatigue, especially<br />

in physical/functioning domains, pain/discomfort, and mobility. These<br />

results indicate that baseline ECOG PS and physician-rated fatigue are<br />

rapid assessments that predict robust measures <strong>of</strong> HRQOL. Future analyses<br />

are planned to examine how HRQOL, ECOG PS and fatigue change over<br />

time with changes in treatment and CLL disease status.<br />

6124 General Poster Session (Board #12B), Mon, 1:15 PM-5:15 PM<br />

Prevalence and predictors <strong>of</strong> adherence to antiemesis prophylaxis in lung<br />

cancer: A population-based study. Presenting Author: Daniel Richard<br />

Gomez, University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX<br />

Background: Nausea/vomiting is a significant toxicity in the treatment <strong>of</strong><br />

lung cancer, but barriers exist to the delivery <strong>of</strong> prophylaxis. We studied<br />

compliance and predictors <strong>of</strong> prophylactic antiemetics with chemotherapy<br />

in this setting. Methods: We used a Texas state registry <strong>of</strong> clinical data<br />

linked with Medicare claims from 2001-2007. Our study population was<br />

incident lung cancers treated with platinum agents within 12 months <strong>of</strong><br />

diagnosis. To define guideline-adherent care, we assessed compliance to<br />

the <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> Oncology recommended prophylactic<br />

agents dexamethasone and a 5-HT3 antagonist. Adherence was scored as a<br />

binary variable and defined as administration within 24 hours <strong>of</strong> the first<br />

day <strong>of</strong> the first cycle <strong>of</strong> chemotherapy. We utilized a logistic regression<br />

model to evaluate the role <strong>of</strong> the following factors in predicting adherence:<br />

concurrent radiation therapy (RT), race (black vs. white), histology (small<br />

cell vs. non-small cell), rural location, Charlson Comorbidity Index (CCI),<br />

household income (by quartiles [Q]), education, and treatment year<br />

(binary). Results: Of 31,762 patients in the database, 5155 patients met<br />

the above criteria. The adherence rate to dexamethasone and 5-HT3<br />

antagonists increased over time, from 51.5% in 2001 to 71.6% in 2007.<br />

Patients treated in the years 2005-2007 were 1.739 times more likely to<br />

be adherent to prophylaxis than were those from 2001-2004. Variables<br />

that predicted adherence (adjusted odds ratio [OR], 95% confidence<br />

interval [CI]) were: age (OR�1.013, CI [1.001, 1.026]), treatment year<br />

(OR�1.756, CI [1.548, 1.991]), race (black vs. white OR�0.684, CI<br />

[0.548,0.853]), median income (higher vs. lower OR�1.83 [Q2 vs. Q1];<br />

OR�1.296 [Q3 vs. Q1]; OR�1.496 [Q4 vs. Q1]), CCI (1� vs. 0,<br />

OR�0.613, CI [0.530,0.709]), and concurrent RT [yes vs. no OR�1.358,<br />

CI [1.197,1.541]). Conclusions: Compliance with guidelines for prophylactic<br />

antiemetics is suboptimal, but increasing over time. Several characteristics<br />

predict for improved adherence, including white race, median income,<br />

advanced age, and the receipt <strong>of</strong> concurrent RT. These findings highlight<br />

substantial economic disparities in supportive care <strong>of</strong> lung cancer in the<br />

state <strong>of</strong> Texas.<br />

6123 General Poster Session (Board #12A), Mon, 1:15 PM-5:15 PM<br />

Loss to follow-up <strong>of</strong> cancer patients in the poorest district <strong>of</strong> the United<br />

States. Presenting Author: Antranik Mangardich, Lincoln Medical and<br />

Mental Health Center, Bronx, NY<br />

Background: Loss to follow-up (LFU) <strong>of</strong> cancer patients is a serious<br />

dilemma, and has only been narrowly studied. Lincoln Medical and Mental<br />

Health Center (LMMHC) serves South Bronx (SB), the poorest district in the<br />

nation. The purpose <strong>of</strong> this study was to assess rates <strong>of</strong> LFU and correlate it<br />

with age, sex, ethnicity, race, cancer types, and stage at diagnosis.<br />

Methods: We collected data from 1,552 patients diagnosed with invasive<br />

cancer in LMMHC between 2006-2010. The data collected were age, sex,<br />

ethnicity, race, type <strong>of</strong> cancer, stage, LFU, treatment, and vital status.<br />

Results: From the 1,552 patients, roughly 25 % were LFU, with 50%<br />

receiving some initial form <strong>of</strong> treatment. The remaining percentages are<br />

shown below (Table). A higher rate <strong>of</strong> LFU was with patients younger than<br />

65 (OR: 1.38, 95% CI: 1.08-1.76). There was no correlation between sex<br />

and LFU. Non-Hispanics were more likely to be LFU compared to Hispanics<br />

(OR: 1.39, 95% CI: 1.07 – 1.8). Blacks were more likely to be LFU<br />

compared to non-Blacks (OR: 1.43, 95% CI: 1.12–1.82). There was no<br />

significance between LFU and stage at diagnosis. Looking at cancer<br />

specific data, colon cancer (C) and head and neck cancers (HN) had the<br />

highest percentage <strong>of</strong> LFU (30% each). There was higher LFU rate for C<br />

compared to breast cancer (B) (OR: 1.7, 95% CI: 1.03-2.8), prostate<br />

cancer (P) (OR: 1.88, 95% CI: 1.18-3.02), and lung cancer (L) (OR: 1.64<br />

95% CI: 0.94- 2.8). HN patients were more likely LFU compared to B (OR:<br />

2.4, 95% CI: 1.13-5.2), P (OR: 2.69, 95 % CI: 1.28-5.68), and L (OR:<br />

2.3, 95% CI: 1.05-5.19) patients. There was no significant difference<br />

between C and HN patients in respect to LFU. Conclusions: In the SB, LFU<br />

rates are related to age, ethnicity, race, and type <strong>of</strong> cancer. Younger<br />

patients, blacks, non-Hispanics, and those with C and HN cancers were<br />

most likely to be LFU, the latter likely due to the lack <strong>of</strong> a HN surgeon at<br />

LMMHC. We hope that with focus on race, ethnicity, and cancer-specific<br />

disparities in LFU rates, we will improve the retention rate <strong>of</strong> our cancer<br />

patients in the future.<br />

Continued to<br />

Sent to nursing<br />

Transferred<br />

LFU<br />

follow up<br />

home/hospice<br />

to other facility<br />

371 (24%) 863 (55%) 167 (11%) 151 (10%)<br />

No Rx Rx No Rx Rx No Rx Rx No Rx Rx<br />

190 (51%)<br />

Rx�treatment<br />

180 (49%) 190 (22%) 672 (78%) 105 (63%) 67 (37%) 93 (62%) 58 (38%)<br />

6125 General Poster Session (Board #12C), Mon, 1:15 PM-5:15 PM<br />

In-hospital mortality <strong>of</strong> patients admitted through the emergency department<br />

<strong>of</strong> a comprehensive cancer center. Presenting Author: Ahmed F.<br />

Elsayem, University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX<br />

Background: Cancer is a common presenting condition for emergency<br />

departments (EDs); however, there is limited information on outcomes <strong>of</strong><br />

ED cancer patients subsequently admitted to the hospital. The purpose <strong>of</strong><br />

this study is to describe outcomes <strong>of</strong> patients with hematologic malignancies<br />

versus those with solid tumors admitted through the ED <strong>of</strong> a<br />

comprehensive cancer center. Methods: We queried the ED database <strong>of</strong> The<br />

University <strong>of</strong> Texas MD Anderson Cancer Center for calendar year 2010 and<br />

linked it to our institutional data warehouse, including tumor registry data.<br />

We classified all leukemia and related disorders, lymphoma, multiple<br />

myeloma, and bone marrow transplant patients as hematologic malignancies,<br />

and remaining cancers as solid tumors. Descriptive statistics, including<br />

chi-square, and t-tests were used in two-sided comparisons. All<br />

statistical analyses were performed using SPSS version 15. Results:<br />

20,732 total ED visits were made by 9,320 unique cancer patients. Of<br />

these, 5,364 (58%) were admitted to the hospital at least once (range 1-13<br />

admits). ED admissions constituted 39% <strong>of</strong> total unique patients admitted<br />

(N�13,753). The main admission indications for solid tumor patients were<br />

infectious complications (particularly pneumonia), intractable pain, or<br />

dehydration. For hematologic malignancies, the main indication was<br />

neutropenic fever. 211/1656 (13%) <strong>of</strong> liquid tumor patients were admitted<br />

to the Intensive Care Unit (ICU) compared to 484/3708 (13%) <strong>of</strong> solid<br />

tumor patients (P�NS). Of all patients admitted through the ED, 587/<br />

5364 (10.9%) died during hospitalization. The hematologic hospital<br />

mortality rate was 225/1653 (13.6%) versus 362/3708 (9.8%) for solid<br />

tumors (P�0.001). Only 242/8389 (3%) <strong>of</strong> patients admitted directly<br />

from outpatient clinics died during the hospitalization (p�0.001).<br />

Conclusions: Patients admitted through the ED, particularly those with<br />

hematologic malignancies, have a high hospital mortality rate. ED-based<br />

palliative care interventions may be justified to improve quality <strong>of</strong> life and<br />

prevent unnecessary costly interventions and ICU admission. Further<br />

research should define predictors <strong>of</strong> poor outcomes in cancer patients<br />

admitted through the ED.<br />

Visit abstract.asco.org and search by abstract for the full list <strong>of</strong> abstract authors and their disclosure information.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!