Jenkins, Sarah M.

227 Background: The objective of this work was to develop a reliable and valid survey measuring aspects of cancer care important to patients to be used in both inpatient and outpatient settings, and that allows comparisons of patient experiences across treatment centers and treatment modalities (medical oncology, radiation, surgery). Other objectives included testing various modes of data collection and developing a standardized sampling methodology based on ICD diagnosis and procedure codes. Methods: We conducted focus groups with patients who had received cancer treatment and with family members of cancer patients. We convened a panel of experts representing quality improvement, oncology, shared decision-making, patient safety, and patient advocacy to obtain feedback on the development of the survey. We interviewed stakeholder groups representing oncology associations to obtain feedback on focal provider, attribution and sampling. We then conducted a field test of the survey among six cancer centers around the country, conducted psychometric analyses, revised the survey, conducted a second field test among four community oncology practices in CA, re-analyzed the data, and finalized the survey. Results: Final survey content was determined using results of the formative research, psychometric analyses, and input from the CAHPS Consortium. The core survey includes 56 questions, 23 of which map to eight composites: Provider Communication, Enabling Patient Self-management, Team Available to Provide Information, Access to Care, Care Coordination, and Office Staff. Supplemental item sets include seven items assessing Shared Decision Making, four items measuring Keeping Patients Informed, and two additional Access to Care items. Mail-only, mail-telephone mixed-mode, and web-mail mixed-mode are recommended methods of data collection. The survey and all recruiting materials are available in English and Spanish. Conclusions: The CAHPS Cancer Care Survey is a rigorously developed, well-tested, reliable and valid survey of patient experiences with their cancer care. The survey and supporting materials are available free of charge on the CAHPS website (www.ahrq.gov/cahps).

Striking cardiovascular health disparities exist among African-Americans in Minnesota compared to Whites; however, community-based interventions to address cardiovascular disease risk are lacking. This study explored participant perceptions of a culturally tailored, cardiovascular disease prevention program developed using a community-based participatory research process. Research participation perceptions, program benefits, and program satisfaction/acceptability were analyzed using a mixed-methods approach. Overall, acceptability was high. Findings highlight the favorable inclusion of African-Americans (research perception), knowledge gained about healthy lifestyle practices (benefits), and quality of the curriculum/speakers (satisfaction). Community-based participatory research may be useful in fostering the acceptability of behavior change interventions among marginalized African-American communities.

86 Background: Phyllodes tumors (PT) are rare breast neoplasms confounding their systematic study and evidence-based management guidelines. There is little data on the sensitivity of preoperative core needle biopsy (CNB) and imaging findings in establishing a correct preoperative diagnosis. We undertook this study to evaluate the sensitivity of CNB histologic findings and imaging findings in preoperatively categorizing tumors as benign or borderline/malignant. Methods: We identified 47 patients who underwent surgical resection of a PT at our institution after a preoperative CNB between 6/2000-3/2012. Statistical analysis utilized Wilcoxon rank-sum, chi-square or Fisher’s exact tests, and 95% confidence intervals (CI) are reported. Results: 30 patients had a final diagnosis of benign and 17 of borderline/malignant PT. The latter were significantly more often palpable (76.5% vs 36.7%, p=0.01). No other clinical or radiologic feature predicted borderline/malignant subtype, although irregular shape on US was suggestive (70.6% vs 44.8%, p= 0.09). CNB diagnosis by tumor type is summarized in the table. No benign PT had ≥10 mitoses, necrosis or marked stromal atypia on CNB. No case with absent mitoses on preoperative CNB was a borderline/malignant PT; however 77% of benign PT did exhibit mitoses on CNB. Sensitivity of CNB for PT overall was 48.9% (95% CI: 35.3-62.8%) while it was 40% (24.6-57.7%) for benign PT and 17.6% (6.2-41.0%) for borderline/malignant PT. Conclusions: Marked stromal atypia, ≥10 mitoses and necrosis were rare on CNB, but suggestive of malignancy. No imaging or histology feature reliably distinguished between tumor types. Over one-third of borderline/malignant PT had a preoperative CNB diagnosis of fibroadenoma or cellular fibroepithelial lesion emphasizing the low sensitivity of CNB and the need for judicious consideration of definitive surgical excision. [Table: see text]

147 Background: The Institute of Medicine recommends development of cancer survivorship care models that are patient centered. We designed a cross-sectional survey to determine the medical and psychosocial needs and preferences of breast cancer survivors. Methods: We identified a random sample, stratified by age and chemotherapy usage, of 600 women from MN, WI, and IA, ages 18-99, within 1-5 years of diagnosis of stage 0-3 breast cancer and seen at our institution. Results: We received 329 surveys (response rate 55%). Respondents were a mean age of 58 years (range 26-89) and a mean of 3.3 years from cancer diagnosis (range 1.2-5.4). 96% were white and stage distribution was: 15% stage 0, 35% stage 1, 32% stage 2, 15% stage 3, and 3% missing. Treatment included mastectomy (60%), lumpectomy (40%), radiation (61%), chemotherapy (55%) and hormonal therapy (50%). Most common treatment-related sequelae were neuropathy (38%), chest wall or arm pain (27%), and lymphedema (26%). Over half (52%) report worry and 45% report fatigue as a problem in the past week. 74% stated that their medical needs were met, whereas 49% said their psychological and spiritual needs were met (p<0.01). 68% understood the follow-up plan after treatment was complete. Perceived barriers to follow up care were anxiety (30%) and lack of coordination (10%). When queried about post-treatment care, the following aspects were considered important (% of patients ranking ≥7 on scale 0-10: 0=not important, 10=very important): breast/chest wall exam (88%); screening for osteoporosis (71%), colorectal cancer (69%), and heart disease (65%); educational materials (64%); mammography (63%); physical therapy (61%); diet and exercise counseling (60%); and complementary/alternative medicine information (53%). Less important were psychology (44%), gynecology (43%), chaplain services (35%), support groups (32%), social work (20%), and infertility experts (4%). 60% expressed interest in attending a survivorship clinic. Conclusions: This survey provides unique data on breast cancer survivors, and efforts should be made to optimize medical and psychosocial care in line with their needs and preferences. These results will be imperative in designing a wellness and survivorship clinic.

Although the majority of giant cell tumors (GCTs) of the bone occur in adult patients, occasionally they arise in the pediatric population. In this setting they may be mistaken for tumors more commonly seen in this age group, including osteosarcoma, aneurysmal bone cyst, and chondroblastoma. All cases of primary GCT of the bone arising in patients 18 years and below were retrieved from our institutional archives and examined with emphasis on the evaluation of various morphologic patterns. Clinical/radiologic records were reviewed when available. Analysis for H3F3A/H3F3B mutations was performed in a subset of cases. Sixty-three (of 710) patients treated at our institution for GCT were 18 years of age and below. The following morphologic patterns were identified: fibrosis (31 cases, 49%), reactive-appearing bone (26, 41%), cystic change (7, 11%), foamy histiocytes (6, 10%), secondary aneurysmal bone cyst (3, 5%), and cartilage (2, 3%). Infarct-like necrosis was present in 17 tumors (27%), and the mitotic rate ranged from 0 to 35 mitoses/10 high-power fields (median 5 mitoses/10 high-power field). Follow-up information (n=55; 6 mo to 69.6 y; median, 11.6 y) showed 21 patients with local recurrence (38%) and 2 patients with lung metastasis (4%). Polymerase chain reaction with sequencing showed that 5 of 5 tested cases harbored H3F3A mutations. In summary, GCT arising in the pediatric population is rare, representing 9% of GCTs seen at our institution. The morphologic spectrum of these tumors is broad and similar to that seen in patients above 18 years of age. It is important to recognize that matrix formation may be observed in GCT, including reactive-appearing bone and cartilage, as well as areas of fibrosis mimicking osteoid production, to avoid misclassification as osteosarcoma or other giant cell–rich lesions common in children.