Ashlee Vennettilli

Patient preferences for research access to administrative data in Ontario.

144 Background: Clinical trials collect outcome data until the trial ends. Subsequent health data is often available in government or health care administrative databases, yet this information is often not made available to researchers. We assessed patient preferences of long-term linkage of such administrative databases to clinical trials databases. METHODS A self-reported questionnaire was administered to outpatients in clinics of academic and community cancer centres. Cancer patients were presented with a hypothetical scenario of a clinical trial and asked about their willingness to allow confidential research access to their personal health information contained in administrative databases. RESULTS Of 524 patients, 54% were females, median age was 60 (range 19-93), 81% were Caucasian, and 61% had post-secondary education. Of cancer sites, 20% had breast cancer, 17% GI, 14% GU, 13% hematology, 12% thoracic, 12% head/neck and 11% gynecologic cancers. One-third (33%) had participated in a clinical trial. An overwhelming 93% allowed confidential access to health information in administrative databases (59% always, 34% depending on circumstances) including access to initials and birth dates, to be used to match information across databases (70% always, 23% depending on circumstances). Only 3% were unwilling to release their information under any circumstances. When asked about other identifiable information, such as name and address, more than 74% allowed the secure storage of this information at the central study coordinating center. In a new cohort of 103 patients, 84% preferred long-term data to be continually available to researchers after the clinical trial has ended; only 9% did not want this option. Results were similar across all sociodemographic subgroups studied including patients who had prior participation in clinical trials, although older patients were even more likely to allow access than younger patients (p = 0.02). CONCLUSIONS The vast majority of patients were willing to have their long-term outcome data collected through government and health care administrative databases made available to clinical trial researchers, to improve our understanding of long-term outcomes of trial procedures and drugs.

Canadian cancer site-specific health utility values: Creating the basis for measuring value and costs of therapy.

7 Background: Health utility values (HUVs) play an integral role when conducting health economic analyses, but a paucity of reference HUVs exists for cancer patients. Using EQ-5D, we generated reference HUVs for multiple malignancies. We further assessed patient willingness to compete the instrument on a regular basis by adding the EQ-5D to an Ontario-wide patient-reported symptom tool mandated by Cancer Care Ontario, the provincial cancer government agency. METHODS 1,831 cancer patients across all non-CNS solid and hematologic cancer sites at the Princess Margaret Cancer Centre completed the EQ-5D instrument; a subset (n=618) were asked about the acceptability of regularly completing the EQ-5D. HUVs were calculated using Canadian valuations. RESULTS The mean±SD HUV for all patients was 0.81±0.15, but were significantly different across different disease sites (p<0.0001): Testicular cancer, 0.87±0.13; prostate, 0.87±0.15; colorectal, 0.83±0.12; head/neck, 0.82±0.15; lymphoma, 0.82±0.15; breast, 0.81±0.17; esophageal, 0.81±0.16; ovarian, 0.79±0.15; leukemia, 0.78±0.15; lung, 0.78±0.13 and myeloma, 0.77±0.14. Confirming the validity of these HUVs, patients with PRO-ECOG scores of 0, 1, 2 and 3 had HUVs of 0.90±0.14, 0.77±0.11, 0.65±0.14 and 0.59±0.19, respectively (p<0.0001). In patients with solid tumors, those with local disease had HUVs of 0.82±0.15; metastatic disease, 0.80±0.15; p=0.015. 88% of patients reported that the EQ-5D was easy to complete, 92% took less than 5 minutes, 89% were satisfied with its length and 86% were satisfied with the types of questions asked. Importantly, 92% reported that they would complete the EQ-5D, even if it was used solely for research purposes and 73% agreed with the notion of completing it regularly at their clinic visits. CONCLUSIONS We present the first Canadian reference dataset of HUVs for common cancers; stage-and site-specific reference values will be presented at the meeting. Mean HUVs varied by disease site, performance status, and disease severity. Furthermore, a majority of patients surveyed were willing to complete the EQ-5D on a regular basis, suggesting that routine administration is feasible across Ontario.

Involving patients to improve their care through real-time patient reported outcome (PRO)-CTCAE chemotoxicity surveys in an outpatient chemodaycare (DC) setting: Evaluating patient acceptability.

64 Background: In a busy DC setting, the efficiency of identifying important treatment toxicities is essential to quality care. Using a systematic approach to collecting patient-reported outcomes in the waiting rooms of DC units is one possible means of improving care while involving patients. This study reports such a pilot study, and the associated assessment of patient acceptance of this approach. METHODS 156 cancer patients over the age of 18 receiving chemotherapy treatment at Princess Margaret Cancer Centre completed a patient-reported chemotoxicity assessment using PRO-CTCAE items on tablet technology. Main symptoms assessed were: fatigue, nausea and vomiting, diarrhea and constipation, pain, aching muscles and/or joints and dysgeusia. Patients perception on the usefulness of PROMs and their willingness to complete such a tool routinely was assessed. RESULTS The median age was 53.5 (range: 19-88 years), 38% were male and 66% were Caucasian. Over 80% did not find the survey overly time consuming (or made their visit more difficult). Less than 1% were distressed by the survey questions. Over 80% reported that the survey asked the appropriate questions. While 81% considered the PROMs useful in informing their physician of their symptoms, 25% reported they would not be willing to complete the survey at each clinic visit. Another 25% were unsure of their feelings toward this approach. 93% were happy to complete the survey using tablet touchscreen technology. CONCLUSIONS Most patients found the survey method of self-reporting ones symptoms to be acceptable, non-distressful, and an important practice. From the patient perspective, the process of reporting ones symptoms using tablet touchscreen technology is both simple and feasible.Yet, only 50% of patients would be willing to complete this survey at every clinic visit. Additional mixed-methods analysis looking at patient characteristics associated with acceptance/non-acceptance and willingness to complete the survey on a regular basis is ongoing and will be reported at the conference.

Improving quality of care by obtaining patient-reported outcomes (PRO)-CTCAE chemotoxicities using tablet technology in daycare (DC) waiting rooms.

165 Background: In a busy chemo DC, any efficient means of tracking important chemotoxicities can improve quality of care. The study goal was to evaluate whether tablet technology available in a DC waiting room is able to capture prevalent and severe toxicities associated with chemotherapy using the patient reported outcome (PRO) - common toxicity criteria for adverse events (CTCAE). METHODS This cross-sectional PRO-CTCAE study of 160 adult solid/hematologic cancer outpatients of all stages, who were undergoing chemotherapy (CT), focused on common chemotoxicities captured using touchscreen tablets in the DC waiting room of Princess Margaret Cancer Centre (Toronto, CA). Individual health scores from the EQ-5D VAS tool and the prevalence of AEs experienced by cancer patients within the past seven days were captured. Symptoms that were listed as moderate to very severe were considered significant. RESULTS Across a wide range of tumours and patients on intravenous CT, the median age (range) was 56 (19-88) years; 38% were males.Patients reported a median (range) health score (100 = best health possible, 0 = worst) of 70 (4-100). The severity offiveprevalent, key side-effects of CT were tabulated (Table). 59% of patients felt their fatigue interfered significantly with their daily activities, and 30% felt decreased appetite interfered significantly. 32% experienced nausea occasionally to almost constantly. CONCLUSIONS The common symptoms of CT were captured FEASIBLY, and found to be highly prevalent in this CT-treated population. Capturing additional symptom prevalence outside of the 7-day time frame may be important from a clinical standpoint. Administration of PRO-CTCAE instrument through tablet technology may be an excellent method to help collect such data systematically and reliably. Updated data on 300 patients will be presented at the meeting. [Table: see text].

The relationship between health utility, quality of life, and symptom scores in Canadian patients with esophageal cancer.

149 Background: Health Utility scores (HUS) are an increasingly important tool in helping to determine the cost-effectiveness of therapies worldwide. The EQ-5D is a validated HUS questionnaire, with reference data in numerous populations. Previously, HUS in esophageal cancer (EC) were based on limited datasets, and the relationship between HUS and either quality of life (QOL, through the validated FACT-E) or esophageal-specific symptoms such as dysphagia, has not been studied. Methods: This cross-sectional survey of EC patients at Princess Margaret Cancer Centre (2012-2014) assessed EQ-5D, FACT-E, a Visual Analog Scale (VAS), patient reported performance status (PRO-ECOG), and dysphagia scoring. EQ-5D scores were converted to HUS using Canadian references. Correlation analyses were performed between HUS and global FACT-E scores, global dysphagia scores, and specific esophageal symptom scores included in FACT-E. Results: Of 198 patients, median age was 67 (range 32-93) years, 76% were male, with localized ...

Quality care and patient involvement in their care: Preferences for transitioning cancer follow-up care to the community.

75 Background: The cancer survivor rate is rising and a new focus has turned to the appropriate survivorship care for this new population. Although several models of survivorship care are currently present in Canada, few studies have examined cancer patient preferences. This study compared patient preferences for direct referral back to ones primary care practitioner (PCP) after 2-3 years of oncology specialist care versustransition through a specialized transitional clinic before exclusive follow-up by ones PCP. The main objectives were to: (1) assess patient interest in a specialized cancer transitional clinic model, and (2) compare patient and demographic characteristics by such preferences. METHODS A cross-sectional scenario-based survey of cancer survivors who had undergone curative therapy at the Princess Margaret Cancer Centre assessed patient preferences for the transitioning of their post-treatment cancer care. Regression models compared clinico-demographic and psychosocial variables (anxiety, depression, distress) to ones preference for transition of care. RESULTS Among 242 cancer survivors, 54% were male, 78% Caucasian, 43% with a college degree, median household income between 60-75K, 71% married, with a wide distribution of curable cancer sites (anal, colorectal, breast, testicular, lymphoma, head/neck, lung). 77% preferred transition through a specialized transitional clinic versus direct referral back to their family doctor. No factors were found to be related to preference for transition of care, except individuals who scored high on anxiety using the ESAS scale were significantly more likely to prefer referral to the specialized transitional clinic (p<0.05). No differences in preference were found by other clinico-demographic factors. CONCLUSIONS A specialized transitional clinic is a preferred survivorship option in three-quarters of cancer patients. Anxiety, but not other variables, was associated with preference for the specialized clinic. This study highlights the need for individual decision-making regarding survivorship options. This individualization of transitioning may help improve patients perception of quality cancer care.

Involving clinic patients in systematic symptom reporting to improve cancer care: Exploring prevalence of sleep disturbances (SD) and fatigue (FAT).

68 Background: SD and FAT occur in 30-50% of cancer patients. Patient-reported outcome measure surveys are avenues through which healthcare providers (HCP) can receive symptom-related clinically relevant information directly from patients, and engage them in their own care plan. By asking patients to report symptoms rapidly through tablet/computer-based technology, HCPs can involve patients in the delivery of care. METHODS In a pilot study evaluating utility of systematic symptom reporting, 336 adult cancer patients across all stages and disease sites who were attending outpatient cancer clinics at Princess Margaret Cancer Centre (PMCC) (Toronto, Canada) completed electronic tablet-administered secure surveys on SD (Insomnia Severity Index) and FAT patterns (FACT-fatigue). These tools measured both symptom severity and interference with function. RESULTS With a median age of 59 (19-91) years, 55% female, across a broad distribution of cancer sites, 56% of our sample reported moderate to very severe (MTVS) SD over the last 7 days: 31% had MTVS difficulty falling asleep; 43% had MTVS difficulties staying asleep; 36% had MTVS problems waking up too early. While 62% who had MTVS SD were not distressed by their SD, 95% who were distressed by their SD met the criteria of MTVS SD. 78% of patients had any level of FAT over the last 7 days, with 40% reporting MTVS FAT. While 40% who had MTVS FAT were still able to perform their usual activities, 67% of patients who were not able to do their usual activities had MTVS FAT. CONCLUSIONS Across all stages and disease sites of cancer patients at PMCC, the prevalence of SD and FAT was both high. Severity and interference with function by FAT and SD were often distinct and non-overlapping. Involving patients in the systematic evaluation of symptoms, particularly using newer tablet-based technology within the clinic, was feasible. Through the use of patient reported electronic applications, patients could easily and systemically report their symptoms in real-time. FAT management has always been a high priority at our institution. However, based on our results, a cancer center-wide self-management plan is being considered for SD.