Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study

Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study J Cancer Surviv (2011) 5:305–314 DOI 10.1007/s11764-011-0173-y Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study Linda C. Harlan & Charles F. Lynch & Theresa H. M. Keegan & Ann S. Hamilton & Xiao-Cheng Wu & Ikuko Kato & Michele M. West & Rosemary D. Cress & Stephen M. Schwartz & Ashley W. Smith & Dennis Deapen & Sonja M. Stringer & Arnold L. Potosky & the AYA HOPE Study Collaborative Group Received: 23 December 2010 /Accepted: 9 January 2011 /Published online: 28 January 2011 # The Author(s) 2011. This article is published with open access at Springerlink.com Abstract population-based sample of AYA survivors to examine Introduction Cancer is rare in adolescents and young adults issues of treatment and health outcomes. (AYA), but these patients have seen little improvement in Methods Individuals diagnosed in 2007–08 and age 15–39 survival in contrast to most other age groups. Furthermore, at the time of diagnosis with acute lymphocytic leukemia, participation in research by AYAs is typically low. We Hodgkin lymphoma, non-Hodgkin lymphoma, germ cell conducted a study to examine the feasibility of recruiting a cancer or sarcoma were identified by 7 Surveillance, The authors have no commercial associations, current and over the past 5 years, that might pose a conflict of interest. L. C. Harlan (*) A. W. Smith I. Kato National Cancer Institute, Karmanos Cancer Institute and Department of Pathology, Wayne State University School of Medicine, 6130 Executive Blvd, MSC 7344, Detroit, MI, USA Bethesda, MD 20892-7344, USA e-mail: lh50w@nih.gov R. D. Cress C. F. Lynch M. M. West Department of Public Health Sciences, Department of Epidemiology, University of Iowa, UC Davis School of Medicine, Iowa City, IA, USA Davis, CA, USA T. H. M. Keegan R. D. Cress Cancer Prevention Institute of California, Public Health Institute/California Cancer Registry, Fremont, CA, USA Sacramento, CA, USA T. H. M. Keegan S. M. Schwartz Department of Health Research and Policy, Public Health Sciences Division, Stanford University School of Medicine, Fred Hutchinson Cancer Research Center, Stanford, CA, USA Seattle, WA, USA A. S. Hamilton D. Deapen Department of Preventive Medicine, Keck School of Medicine, S. M. Stringer University of Southern California, Kelly Services, Scientific Resources, Los Angeles, CA, USA 6101 Executive Boulevard, Suite 392, Rockville, MD, USA X.-C. Wu Epidemiology Program/Louisiana Tumor Registry, School of A. L. Potosky Public Health, Louisiana State University Health Sciences Center, Lombardi Comprehensive Cancer Center, Georgetown University, New Orleans, LA, USA Washington, DC, USA 306 J Cancer Surviv (2011) 5:305–314 Epidemiology, and End-Results (SEER) cancer registries, health-related quality of life, unmet needs, and the impact of mailed surveys within 14 months after diagnosis and again cancer on psychosocial domains. The design and implemen- a year later, and had medical records reviewed. tation of the Adolescent and Young Adult Health Outcomes Results 525 (43%) of the eligible patients responded, 39% and Patient Experience (AYA HOPE) Study was motivated refused and 17% were lost to follow-up. Extensive efforts and guided by the findings of a systematic review of the were required for most potential respondents (87%). 76% science on AYA cancer patients [5]. of respondents completed the paper rather than online Here we report on the feasibility of 1) recruiting AYA survey version. In a multivariate model, age, cancer site, cancer survivors, 2) developing and fielding a patient education and months from diagnosis to the first mailing survey with a subsequent survey to examine changes over of the survey were not associated with participation, time and 3) obtaining patient (or guardian) consent to although males (p<0.01), Hispanics and non-Hispanic review medical records. We focus on the practical chal- blacks (p<0.001) were less likely to participate. 91% of lenges in recruiting and conducting research in AYA cancer survivors completing the initial survey completed the survivors identified through population-based registries subsequent survey. across the United States, and discuss potential strategies Discussion Despite the response rate, those who participated to increase recruitment of these survivors. adequately reflected the population of AYA cancer survivors. The study demonstrates that cancer registries are valuable foundations for conducting observational, longitudinal Methods population-based research on AYA cancer survivors. Implications for Cancer Survivors Achieving a reasonable Study setting and eligibility criteria response rate in this population is possible, but requires extensive resources. Patients were identified through seven population-based Surveillance, Epidemiology and End-Results (SEER) pro- . . Keywords Adolescent cancer Young adult cancer gram cancer registries: Detroit; Seattle/Puget Sound; Los . . . Survey Response rates Medical records Consent forms Angeles County, San Francisco/Oakland, Greater California (13 counties around Sacramento plus Orange County), and the states of Iowa and Louisiana. Introduction Conduct of this study required approval from 9 Institu- tional Review Boards (IRBs) (7 registries, 1 state (California) Although cancer is rare in adolescents and young adults and the NCI IRB), which took 7 months. Several registries (AYA), it is the leading cause of death in this age group experienced particularly long approval processes because the after unintentional injury, homicide and suicide [1]. inclusion of minors required changes to consent forms. No Survival following a cancer diagnosis in AYAs (defined as concerns were raised by the IRB about survey content, ages 15–39) has shown little improvement relative to other although there was concern about security of the online age groups [2]. In particular, cancer patients diagnosed survey that was resolved with a more complete explanation of between the ages of 25 and 35 have had no survival the security protocols. The online survey website included a improvement in more than 20 years [2]. Although scientific firewall, Virtual Private Network (VPN), an intrusion evidence to understand reasons for this phenomenon is detection system, and routine security checks of the computer limited, experts have speculated that the lack of improve- resources. Patients were given a website address, user name ment in survival may be due to a combination of factors, and password in the initial mailing. Patients accessing the such as the lack of available clinical trials hampering efforts online survey were required to create a new password upon to develop novel therapies,[3] lack of medical insurance, entering the system. This process made the survey accessible poor access to medical care for the initial diagnosis and for only to the person with the new password. Once the survey follow-up care, attitudes of invincibility, physicians’ low was submitted no further access was allowed. suspicion of cancer resulting in delayed diagnosis, and Eligible patients were diagnosed between July 1, 2007 location and specialty of treating physicians [4]. and October 31, 2008 with a first invasive, histologically To address several of these issues in AYA cancer patients, confirmed germ cell cancer, non-Hodgkin lymphoma the National Cancer Institute (NCI) in 2007, in partnership (NHL), Hodgkin lymphoma (HL), acute lymphocytic with the Lance Armstrong Foundation (LAF), initiated an leukemia, Ewing’s sarcoma, osteosarcoma, or rhabdomyo- observational cohort study to: 1) examine factors associated sarcoma (Appendix I), ages 15 through 39 years at with high quality cancer treatment in general community diagnosis, residents of the study area and able to read practice, including the use of clinical trials and treatment English. The goal was to obtain 530 surveys completed 6– protocols, and 2) assess patient-reported outcomes, such as 14 months from the date of diagnosis. A small sample of J Cancer Surviv (2011) 5:305–314 307 deceased patients who were otherwise eligible was included website directories (e.g., Google, Zabasearch), paid direc- in the medical record review. tory services (e.g., Accurint, Coles) and/or public records (e.g., Department of Motor Vehicles, Voter Registration). Data collection instruments A few study sites sought new addresses by exploring social networking websites. If a new address was not The survey (http://outcomes.cancer.gov/surveys/aya/), found, the staff attempted to contact the individuals by which took approximately 20 min to complete, was telephone. If the interviewer received an answering designed to be self-administered and address a number machine a message was left; after leaving 2 messages no of issues including the impact of cancer, health-related additional messages were left. Phone calls were made at quality of life, healthcare delivery, and reasons for not different times during the day, evening and weekends. If participating in clinical trials. Survey development included none of these efforts was successful, the patient was cognitive and usability testing with 24 AYA cancer survivor classified as lost to follow-up. volunteers. Patients who completed the survey were sent a thank Patients were asked to complete a healthcare provider you letter and $25.00 for the time spent completing it. form by listing the names and addresses of all healthcare An additional $25.00 was provided for time spent providers and facilities providing care. The information was completing the healthcare provider and medical record used to supplement registry information and to request release forms. medical records for data abstraction. Medical record retrieval Patient recruitment Medical records were collected from facilities listed on the Before patients were contacted, the patient’s physician was healthcare provider form. Information obtained included notified that his/her patient would be approached to type of healthcare facilities, physician subspecialties, tumor participate in the study. Physicians were asked to advise characteristics and staging, diagnostic procedures, partici- study staff if there were reasons the patient should not be pation in active clinical trials, therapy provided, and contacted. If the staff had not heard from the physician in comorbid conditions. 3 weeks, they contacted the patient. Some physicians and one state registry required active physician consent. Follow-up survey Eligible patients were mailed a cover letter, a brochure introducing the study, an overall study consent document We also conducted a follow-up survey, designed to be self- administered by paper or online, to examine changes in (where required by the registry’s IRB), healthcare provider form, a Health Insurance Portability and Accountability Act psychosocial outcomes and health-related quality of life (HIPAA) compliant consent for the release of medical 15–35 months following the initial diagnosis. Survey information, a paper survey, a web address for the online development (http://outcomes.cancer.gov/surveys/aya/), survey, a pre-addressed stamped return envelope for study tracing and follow-up was similar to the methods used for materials, and a LIVESTRONG bracelet. If the patient was the initial survey. Participants were provided with $50 for a minor at the time of contact, study materials were mailed time spent completing the survey. No additional medical to the parent/guardian and included a consent document record review was performed. requiring signatures from the parent/guardian and a signa- ture of assent from the minor. Analyses Follow-up procedures for non-respondents included a second mailing 3 weeks after the initial mailing, followed Descriptive univariate analyses were conducted to com- by telephone calls from trained interviewers beginning pare demographic and tumor characteristics of baseline 2 weeks after the second mailing. If needed, multiple survey respondents and non-respondents based on data follow-up calls were made to ask the patient to complete collected by each SEER registry. We used the patient’s the survey over the telephone and mail the medical record address at the time of cancer diagnosis to determine area release and the healthcare provider form. level educational attainment and median family income at If study materials were returned marked undeliverable, the census tract level [6]. A multiple logistic regression the staff employed a variety of tracing procedures to obtain model was used to assess the association between clinical and the correct address, including reviewing the cancer registry non-clinical factors and participation. Survey participants database for updated information, contacting directory were compared by survey mode, and follow-up survey assistance, obtaining information from healthcare providers, completion. A significance level of 0.05 (two tailed) was hospital(s) and family members. Study sites also used used for all analyses. 308 J Cancer Surviv (2011) 5:305–314 Results remaining 83 patients, neither a valid telephone number nor an address could be found. A total of 1,405 patients were identified as potentially eligible (Fig. 1). Among registries that required physician Generalizability of the enrolled sample consent, only 6 physicians refused to allow their patient to be contacted. Physicians identified 11 patients as We compared the characteristics of patients who participated deceased and 9 as ineligible. Another 70 patients became in the survey to non-responders (Table 1). In univariates ineligible because they exceeded their 14-month eligibil- analyses, non-respondents were not significantly different ity date prior to contact. After mailing the survey (n= from responders by age, census tract education, or census 1,309), we identified 16 deceased patients and 85 patients tract median family income. Response did vary by cancer ineligible because they did not speak English (56%), site (p<0.04) from 38% for acute lymphocytic leukemia and they denied having cancer (9%), or due to other reasons sarcoma to 51% for HL. Females were more likely to (34%). participate (p<0.0001). Non-Hispanic black (p<0.05) and Hispanic (p<0.001) patients were less likely to respond than Response rates non-Hispanic whites. Among non-respondents, non- Hispanics whites and non-Hispanic Blacks were more likely A total of 524 eligible patients completed the initial survey to refuse (p<0.001) while Hispanics (32%) were more than and 1 patient completed only the medical record release twice as likely to be lost to follow-up as non-Hispanic white (Fig. 1), yielding a response rate of 43%. Of patients sent (12%) or non-Hispanic black (14%) patients (data not the initial mailing, 87% required additional contact. shown). Of the refusals, 133 directly declined participation when In a multivariate logistic regression model, we investi- contacted and 341 patients never completed the surveys gated factors thought to be associated with response rate despite repeated contacts (Fig. 1). Of patients lost to follow- (Table 2). Females were more likely to answer the survey up, telephone numbers were found for 126, but several calls than were males (p <0.001). Compared to non-Hispanic to these numbers did not produce a mailing address. For the whites, Hispanics and non-Hispanic blacks were less likely Fig. 1 Patient number AYA HOPE PATIENT RECRUITMENT FLOWCHART flowsheet for the study Deceased Potential Eligible Patients Physician Consent N=11 N=1405 Refused N=6 Ineligible N=79 Deceased Refusals Potential Participants Mailed N=16 N=474 Survey N=1309 Ineligible Lost to Follow-up N=85 N=209 Total Eligible Surveys Completed N=524 Medical Record Consent only Lost Survey N=1 N=1 Paper Web Telephone N=397 N=115 N=12 Medical Record Medical Record Consent Refusals Consents N=26 N=499 Medical Records Obtained N=490 J Cancer Surviv (2011) 5:305–314 309 Table 1 Percent distribution of demographic characteristics for respondents and non-respondents ALL Germ cell HL NHL Sarcoma All cancer sites p-value R NR R NR R NR R NR R NR R NR (n=21) (n=35) (n=206) (n=304) (n=142) (n=136) (n=131) (n=167) (n=25) (n=41) (n=525) (n=683) Age (years) p=0.0788 15–19 57.1 42.9 7.8 12.5 12.7 20.6 9.9 9.6 32.0 56.1 12.8 17.6 20–24 14.3 28.6 21.4 18.4 16.9 22.1 12.2 14.4 24.0 17.1 17.7 18.6 25–29 14.3 11.4 27.2 24.0 31.7 19.1 19.1 18.6 20.0 9.8 25.5 20.2 30–34 0 8.6 24.3 24.0 21.1 18.4 25.2 24.0 12.0 12.2 22.1 21.4 35–39 14.3 8.6 19.4 21.1 17.6 19.9 33.6 33.5 12.0 4.9 21.9 22.3 SEER Race/ethnicity p=0.0026 Hispanic 33.3 31.4 22.3 33.2 12.0 11.8 17.6 21.6 16.0 29.3 18.5 25.8 Non-Hispanic White 47.6 51.4 70.4 53.6 62.7 65.4 59.5 44.3 40.0 41.5 63.2 52.9 Non-Hispanic Black 9.5 2.9 2.4 4.6 12.0 15.4 10.7 16.8 16.0 12.2 8.0 10.1 Other/Unknown 9.5 14.3 4.9 8.6 13.4 7.4 12.2 17.4 28.0 17.1 10.3 11.3 Sex p<0.0001 Male 47.6 65.7 90.3 90.8 39.4 56.6 49.6 60.5 64.0 68.3 63.4 73.9 Female 52.4 34.3 9.7 9.2 60.6 43.4 50.4 39.5 36.0 31.7 36.6 26.1 Census: Median% HS Education (age 25 or older) p=0.0945 Not Available 0 0 0.5 0 0 1.5 0 0 0 0 0.2 0.3 <75% 23.8 25.7 28.2 33.2 26.1 24.3 26.0 28.1 24.0 41.5 26.7 30.3 75–90% 52.4 45.7 37.4 43.4 41.5 41.9 35.1 38.3 52.0 29.3 39.2 41.1 90+% 23.8 28.6 34.0 23.4 32.4 32.4 38.9 33.5 24.0 29.3 33.9 28.3 Census: median family income p=0.1102 Not available 0 0 0.5 0 0 1.5 0 0 0 0 0.2 0.3 <$45,000 19.0 34.3 31.1 38.5 28.9 28.7 29.0 33.5 40.0 39.0 29.9 35.1 $45–$65,000 47.6 34.3 36.9 34.2 36.6 30.9 38.9 32.9 32.0 26.8 37.5 32.8 $65,000+ 33.3 31.4 31.6 27.3 34.5 39.0 32.1 33.5 28.0 34.1 32.4 31.8 R respondent; NR non-respondent; ALL acute lymphocytic leukemia; HL Hodgkin lymphoma; NHL non-Hodgkin lymphoma p-value=0.0372 to respond. Other factors in Table 2 were not significantly Medical records were obtained for 93% of all respondents. associated with participation. However, we were unable to obtain medical records for 2% of consenting participants because the records were Response by survey mode lost within the healthcare facility or the facility failed to provide them after multiple requests. Medical records Some patients chose to complete the survey online (22%) were obtained for a small sample (n=27) of deceased or over the telephone (2%) (Fig. 1 and Table 3). Compared patients primarily to assess any differences in the diagnosis to patients in other age or racial/ethnic groups, fewer 20– and treatment that might arise from excluding patients who 24 year-olds or non-Hispanic black patients and less than 10% had died. of sarcoma patients completed the survey online. Patients with an Associate Degree or higher or who lived in higher Follow-up survey socioeconomic status census tracts, defined by education and income, were more likely to complete the survey online. Of the 524 eligible participants who completed the initial survey, 465 completed the follow-up survey, 34 were non- Medical records responders, and 10 were lost to follow-up (Table 4). The participation rate in the follow-up survey was over 91% Among participants completing the survey, 5% refused to among survivors. Between the initial and subsequent sign a medical records release consent form (Fig. 1). survey (8–17 months), 3% (n=15) of the patients died. 310 J Cancer Surviv (2011) 5:305–314 Table 2 Multivariable analysis examining characteristics associated one recent health survey in adolescents ages 13–17 years with response vs. non-response to the AYA HOPE Study Baseline Survey yielded a 40% response rate [8]. These adolescents had been seen within the healthcare system in the previous year 95% Confidence Interval and were recruited by that system. A national study of childhood cancer survivors between 1994 and 2000 Variable OR Lower Upper p-value achieved a 62% completion rate with extensive resources [9]. However, the denominator for the completion rate Age at diagnosis (years) 0.06 included only those subjects who expressed an interest in 15–19 1.0 the study after being contacted by their treating institution. 20–24 1.3 0.9 2.0 As indicated by our follow-up survey, once individuals 25–29 1.8 1.2 2.7 agree to participate in a study, they are more likely to 30–34 1.5 0.97 2.2 complete subsequent surveys. The response rate in our 35–39 1.4 0.9 2.0 study might have been higher if, as in the NHL [7] study, SEER Racial/ethnicity 0.01 more time had elapsed since diagnosis. Some of the patients Non-Hispanic White 1.0 eligible for our study who refused may still have been Hispanic 0.6 0.5 0.9 under active treatment and may have agreed to participate Non-Hispanic Black 0.6 0.4 0.9 once therapy was completed. Other/Unknown 0.7 0.5 1.0 Males and non-Hispanic black and Hispanic patients SEER Sex <0.001 were significantly less likely to participate, similar to Male 1.0 differences reported for cancer clinical trials [10, 11]. Female 1.7 1.3 2.2 Significantly lower enrollment for minority compared to Cancer site 0.60 white patients and for men compared to women was Acute lymphocytic leukemia 1.0 reported for surgical trials [11]. This finding has been Germ cell cancer 1.1 0.6 2.0 shown not only in cancer, but in other diseases as well [12]. Hodgkin lymphoma 1.4 0.7 2.5 Although 37% of our respondents were racial/ethnic Non-Hodgkin lymphoma 1.1 0.6 2.0 minorities, it is likely to have been higher if translated Sarcoma 1.1 0.5 2.3 versions of the survey were available, as not speaking Census:% HS education 0.31 English was the primary reason for ineligibility among our (age 25 or older) population. <75% 1.0 A variety of methods were used to maximize responses 75–90% 0.9 0.7 1.2 to the mailing. We enclosed a LIVESTRONG bracelet that 90%+ 1.1 0.8 1.6 added interest and bulk to the packet. One registry initially Months from diagnosis to 1.0 0.9 1.01 0.14 used an overnight delivery service, but abandoned this initial mailing of survey within a month. Response rates were lower than with OR>1.0 indicates more likely to respond; OR<1.0 indicates less USPS. The overnight service did not leave packages if no likely to respond. one was home, had limited re-delivery, no forwarding addresses, and deliveries were limited to weekdays. While Discussion we used a variety of approaches to enhance our response rates, we did not send pre-notification letters or postcards. The goal of the AYA HOPE Study was to learn about Research has suggested this may not increase the response recruitment of AYAs, their cancer care and outcomes and to rate [12], but it would have identified incorrect addresses at determine the feasibility of collecting survey and detailed a lower cost than mailing the complete packet. medical record information on a representative population- Overall, 87% of the identified sample required at least based sample of AYA cancer survivors. Our response rate one additional mailing or telephone contact. An address or among eligible patients was 43%. This age group, partic- telephone number could not be found for 16% of patients ularly those ages 15–25 years, is mobile and difficult to identified as eligible, similar to the 15% reported by the follow due to educational and employment opportunities, childhood cancer survivors study [9]. Although we likely marriage and other personal life changes. had the correct phone number for some patients, no one Previous studies have shown that response rates of answered our calls. With the increased use of caller young adults are lower than for older adults [7–9]. A survey identification, patients may have been screening calls from of adult NHL survivors identified from a cancer registry unrecognized phone numbers. It is possible that calls had a 55% response rate, but patients were older and 2– originating from patients’ own medical institutions may 5 years post diagnosis [7]. Despite aggressive follow-up, have yielded better responses, although a response rate of J Cancer Surviv (2011) 5:305–314 311 Table 3 Mode of survey Paper Online Phone completion by patient characteristics, AYA HOPE N % N% N% study Age at diagnosis (years) 15–19 51 76.1% 14 20.9% 2 3.0% 20–24 75 80.6% 14 15.1% 4 4.3% 25–29 99 73.9% 31 23.1% 4 3.0% 30–34 84 73.0% 29 25.2% 2 1.7% 35–39 88 76.5% 27 23.5% 0 SEER Race/ethnicity Hispanic 75 78.1% 19 19.8% 2 2.1% Non-Hispanic White 252 75.9% 72 21.7% 8 2.4% Non-Hispanic Black 34 81.0% 6 14.3% 2 4.8% Other/Unknown 36 66.7% 18 33.3% 0 SEER Sex Male 240 72.3% 82 24.7% 10 3.0% Female 157 81.8% 33 17.2% 2 1.0% Cancer site ALL 18 85.7% 3 14.3% 0 NHL 95 72.5% 32 24.4% 4 3.1% HL 108 76.1% 32 22.5% 2 1.4% Germ cell 153 74.6% 46 22.4% 6 2.9% Sarcoma 23 92.0% 2 8.0% 0 Education Less than high school diploma 53 91.3% 4 6.9% 1 1.7% Completed high school 69 75.8% 16 17.6% 6 6.6% Some college/vocational/training 106 76.8% 30 21.7% 2 1.4% Associate Degree 38 71.7% 15 28.3% 0 College graduate 94 71.8% 34 26.0% 3 2.3% Post-graduate education 35 68.6% 16 31.4% 0 Missing 2 100.0% 0 0 Census:% HS Education (age 25 or older) Not Available 1 100.0% 0 0 <75% 109 78.4% 24 17.3% 6 4.3% 75–90% 166 80.6% 35 17.0% 5 2.4% 90%+ 121 68.0% 56 31.5% 1 0.6% Census: median family income Not Available 1 100.0% 0 0 <$45,000 127 81.4% 22 14.1% 7 4.5% $45,000–$65,000 149 75.6% 45 22.8% 3 1.5% $65,000+ 120 70.6% 48 28.2% 2 1.2% SEER Registry Northern California 81 65.9% 38 30.9% 4 3.3% Los Angeles 77 81.1% 17 17.9% 1 1.1% Seattle-Puget Sound 63 74.1% 22 25.9% 0 Louisiana 52 74.3% 12 17.1% 6 8.6% a Greater California 43 72.9% 16 27.1% 0 ALL acute lymphocytic leukemia; HL Hodgkin Metropolitan Detroit 44 84.6% 7 13.5% 1 1.9% lymphoma; NHL non-Hodgkin Iowa 37 92.5% 3 7.5% 0 lymphoma 312 J Cancer Surviv (2011) 5:305–314 Table 4 Follow-up survey Survey Completed Non-Response Deceased Lost to follow-up completion by patient characteristics, AYA HOPE N% N % N % N % Study Age at diagnosis (years) 15–19 63 94.0% 1 1.5% 0 3 4.5% 20–24 81 87.1% 8 8.6% 2 2.2% 2 2.2% 25–29 114 85.1% 13 9.7% 5 3.7% 2 1.5% 30–34 103 89.6% 6 5.2% 5 4.3% 2 1.7% 35–39 104 90.4% 7 6.1% 3 2.6% 1 0.9% SEER Race/ethnicity Hispanic 83 86.5% 6 6.2% 3 3.1% 5 5.2% Non-Hispanic White 298 89.8% 22 6.6% 8 2.4% 4 1.2% Non-Hispanic Black 35 83.3% 4 9.5% 2 4.8% 1 2.4% Other/Unknown 49 90.7% 3 5.6% 2 3.7% 0 SEER Sex Male 287 86.4% 26 7.8% 11 3.3% 9 2.7% Female 178 92.7% 9 4.7% 4 2.1% 1 0.5% Cancer site ALL 18 85.7% 1 4.8% 1 4.8% 1 4.8% NHL 113 86.3% 9 6.9% 7 5.3% 2 1.5% HL 130 91.5% 10 7.0% 1 0.7% 1 0.7% Germ cell 181 88.3% 15 7.3% 4 2.0% 6 2.9% Sarcoma 23 92.0% 0 2 8.0% 0 Education Less than high school diploma 51 87.9% 3 5.2% 3 5.2% 1 1.7% Completed high school 79 86.8% 8 8.8% 1 1.1% 3 3.3% Some college/vocational training 123 89.1% 8 5.8% 4 2.9% 3 2.2% Associate degree 47 88.7% 5 9.4% 1 1.9% 0 College graduate 119 90.8% 7 5.3% 4 3.1% 1 0.8% Post-graduate education 45 88.2% 3 5.9% 2 3.9% 1 2.0% Missing 1 50.0% 0 0 1 50.0% Census:% HS education (age 25 or older) Not Available 1 100.0% 0 0 0 <75% 121 87.1% 9 6.4% 5 3.6% 5 3.6% 75–90% 182 88.3% 14 6.8% 7 3.4% 3 1.5% 90%+ 161 90.4% 12 6.7% 3 1.7% 2 1.1% Census: median family income Not Available 1 100.0% 0 0 0 <$45,000 131 84.0% 13 8.3% 6 3.8% 7 4.5% $45,000–$65,000 177 89.8% 14 7.1% 5 2.5% 1 0.5% $65,000+ 156 91.8% 8 4.7% 4 2.4% 2 1.2% SEER Registry Northern California 108 87.8% 9 7.3% 2 1.6% 4 3.3% Los Angeles 81 85.3% 7 7.3% 4 4.2% 4 4.2% Seattle-Puget Sound 76 89.4% 6 7.1% 3 3.5% 0 Louisiana 60 85.7% 6 8.6% 4 5.7% 0 a Greater California 57 96.6% 0 1 1.7% 1 1.7% ALL acute lymphocytic leukemia; HL Hodgkin Metropolitan Detroit 49 94.2% 1 1.9% 1 1.9% 1 1.9% lymphoma; NHL non-Hodgkin Iowa 34 85.0% 6 15.0% 0 0 lymphoma J Cancer Surviv (2011) 5:305–314 313 only 40% was reported by a study conducted within dinal population-based research on younger, non-pediatric patients in a healthcare organization [8]. cancer survivors. Some survivors refused to participate in any part of our study because of the HIPAA medical release form. Open Access This article is distributed under the terms of the Evidence from a community-based, randomized, mixed- Creative Commons Attribution Noncommercial License which per- mits any noncommercial use, distribution, and reproduction in any mode survey (n=6,939) reported the inclusion of a medium, provided the original author(s) and source are credited. minimally burdensome version of HIPAA authorization form reduced survey response rates by up to 15% [13]. Simply requiring a signature reduces the response rate. Appendix I Nelson reported that response rates were 58% at locations requiring no advanced permission to contact the individual Sites and tumor morphologies (ICD-O-3 codes) included: compared to 27% for those requiring written permission from the individual [14]. non-Hodgkin lymphoma (9590–9596, 9670, 9671, Although some staff used social networking websites to 9673, 9675, 9678–9680, 9684, 9687, 9689–9691, find addresses, we did not use this approach to communicate 9695, 9698–9702, 9705, 9708, 9709, 9714–9719, with patients because registries did not request IRB approval 9727–9729); for this type of contact. We were concerned about the Hodgkin lymphoma (9650–9655, 9659, 9663–9665, confidentiality of a social network contact and identification 9667); of the correct person. However, in future studies, social germ cell cancer (9060–9091, 9100–9102) excluding networking websites may be a useful method of contact [15]. patients with intracranial and intraspinal tumors (C700- This may increase participation as email/web communication C729, C751-C753); is used extensively among young people. However, methods acute lymphocytic leukemia (9820, 9832–9837); to ensure patient confidentiality are required. Ewing’s sarcoma (9260, 9365), osteosarcoma (9180– The majority of patients in our study completed the 9187, 9192–9195) and rhabdomyosarcoma (8900–8921): paper rather than the online version. Future studies might Sarcomas arising in the central nervous system (C700- evaluate different approaches to increase participation and C729) were excluded. the use of online surveys. A web address included in the initial mailing that links to a well-designed website describing the study and reasons to participate, and includes References the survey as well as other survivorship information might entice AYAs to participate. However, it is possible that 1. Heron MP. Deaths: Leading causes for 2004. National vital AYAs simply prefer to complete the paper version of the statistics reports: vol 56 no 5. Hyattsville, MD: National Center survey. A study in Olmstead County found the inclusion of for Health Statistics, 2007. 2. Bleyer A, Budd T, Montello M. Adolescents and young adults an internet option decreased the response rate [16] while a with cancer: the scope of the problem and criticality of clinical study in Norway found no increase with the online option trials. Cancer. 2006;107(7 Suppl):1645–55. [17]. One possibility is that unless thrown away, a paper 3. Ferrari A, Montello M, Budd T, Bleyer A. The challenges of survey is a constant reminder, whereas a computer can be clinical trials for adolescent and young adults with cancer. Pediatr Blood Cancer. 2008;50:1101–04. turned off. 4. Loberiza FR, Cannon AJ, Weisenburger DD, Vose JM, Moehr MJ, Bast MA, et al. Survival disparities in patients with lymphoma according to place of residence and treatment provider: a Conclusion population-based study. J Clin Oncol. 2009;27:5376–82. 5. Albritton K, Caligiuri M, Anderson B, Nichols C, Ulman D, Adams H, et al. Closing the Gap: Research and Care Imperatives The AYA HOPE Study demonstrates that recruiting and for Adolescents and Young Adults with Cancer: Report of the following a diverse population of AYA survivors diagnosed Adolescent and Young Adult Oncology Progress Review Group. US DHHS, NIH, NCI, LIVESTRONG Young Adult Alliance, with different cancers and living throughout the US can be 108 pages [report on internet] available from: http://women. accomplished using population-based registries. However, cancer.gov/library/AYAO_PRG_Report_2006_FINAL.pdf. it is important for researchers to be realistic in their 6. http://www.census.gov/main/www/cen2000.html. expectations, recognizing that this population is mobile 7. Bellizzi KM, Miller MF, Arora NK, Rowland JH. Positive and negative life changes experienced by survivors of non-Hodgkin’s and difficult to contact. Achieving a high response rate is lymphoma. Ann Behav Med. 2007;34:188–99. challenging, requiring more extensive resources for follow- 8. Richards J, Wiese C, Katon W, Rockhill C, McCarty C, Grossman up. Despite these limitations and challenges, the AYA D, et al. Surveying Adolescents enrolled in a regional health care HOPE Study confirms that the use of cancer registries is a delivery organization: mail and phone follow-up—What works at valuable foundation for conducting observational, longitu- what cost? JABFM. 2010;23:534–41. 314 J Cancer Surviv (2011) 5:305–314 9. Robison LL, Mertens AC, Boice JD, Breslow NE, Donaldson SS, 17. Brogger J, Nystad W, Cappelen I, Bakke P. No increase in Green DM, et al. Study design and cohort characteristics of the response rates by adding a web response option to a postal childhood cancer survivor study: a multi-institutional collabora- population survey: A randomized trial. J Med Internet Res. 2007;9 tive project. Med Pediatr Oncol. 2002;38:229–39. (5):e40. Available from: http://www.jmir.org/2007/e40/. 10. Bleyer A, Budd T, Montello M. Older adolescents and young adults with cancer and clinical trials: Lack of participation and progress in North America. In: Bleyer A, Barr R, editors. Cancer in adolescent and young adults. Heidelburg: Springer Verlag; 2007. p. 71–81. AYA HOPE study collaborative group 11. Stewart JH, Bertoni AG, Staten JL, Levine EA, Gross CP. Participation in surgical oncology clinical trials: gender-, race/ethnicity-, and age- based disparities. Ann Surg Oncol. 2008;15:1818–9. California Cancer Registry/Public Health Institute (Sacramento, 12. Beebe TJ, Rey E, Ziegenfuss JY, Jenkins S, Lackore K, Talley CA): Rosemary Cress, DrPH (P.I.); Gretchen Agha; Mark Cruz NJ, et al. Shortening a survey and using alternative forms of Fred Hutchinson Cancer Research Center (Seattle, WA): Stephen prenotification: impact on response rate and quality. BMC Med M. Schwartz, Ph.D. (P.I.); Martha Shellenberger; Tiffany Janes Res Methodol 2010; 10:50. Available from: http://www. Karmanos Cancer Center (Detroit, MI): Ikuko Kato, Ph.D (P.I.); biomedcentral.com/1471-2288/10/50. Ann Bankowski; Marjorie Stock 13. Beebe TJ, Talley NJ, Camilleri M, Jankins SM, Anderson KJ, Louisiana State University (New Orleans, LA): Xiao-cheng Wu, Locke GR. The HIPAA authorization form and effects on survey M.D., MPH (P.I.); Vivien Chen, Ph.D.; Bradley Tompkins response rates, non-response bias, and data quality: a randomized Cancer Prevention Institute of California (Fremont, CA): Theresa community study. Med Care. 2007;45:959–65. Keegan, Ph.D, M.S. (P.I.); Laura Allen; Zinnia Loya; Karen Hussain 14. Nelson K, Elena Garcia R, Brown J, Mangione CM, Louis TA, University of Iowa (Iowa City, IA): Charles F. Lynch M.D., Ph.D. Keller E, et al. Do patient consent procedures affect participation (P.I.); Michele M. West, Ph.D.; Lori A. Odle, R.N. rates in health services research? Med Care. 2002;40:283–8. University of Southern California (Los Angeles, CA): Ann 15. Lenhart A, Purcell K, Smith A, Zickuhr K. Social media & mobile Hamilton, Ph.D (P.I.); Jennifer Zelaya; Mary Lo; Urduja Trinidad Internet use among teens and young adults. In Pew Internet & National Cancer Institute (Bethesda, MD): Linda C. Harlan, BSN, American Life Project. Pew Research Center: Washington, DC; 2010. MPH, Ph.D.; (Investigator) Ashley Wilder Smith, Ph.D, MPH (Co- 16. Ziegenfuss JU, Beebe TJ, Rey E, Schleck C, Locke GR, Talley investigator); Sonja M. Stringer, MPH; Gretchen Keel, BS,BA. NJ. Internet option in a mail survey: more harm than Good? Consultants: Arnold Potosky, PhD.; Keith Bellizzi, Ph.D.; Karen Epidemiology. 2010;21:585–86. Albritton, MD, Michael Link, MD; Brad Zebrack, Ph.D., http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Cancer Survivorship Pubmed Central

Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study

Loading next page...
 
/lp/pubmed-central/recruitment-and-follow-up-of-adolescent-and-young-adult-cancer-ffYZCNyHIX

References (30)

Publisher
Pubmed Central
Copyright
© The Author(s) 2011
ISSN
1932-2259
eISSN
1932-2267
DOI
10.1007/s11764-011-0173-y
Publisher site
See Article on Publisher Site

Abstract

J Cancer Surviv (2011) 5:305–314 DOI 10.1007/s11764-011-0173-y Recruitment and follow-up of adolescent and young adult cancer survivors: the AYA HOPE Study Linda C. Harlan & Charles F. Lynch & Theresa H. M. Keegan & Ann S. Hamilton & Xiao-Cheng Wu & Ikuko Kato & Michele M. West & Rosemary D. Cress & Stephen M. Schwartz & Ashley W. Smith & Dennis Deapen & Sonja M. Stringer & Arnold L. Potosky & the AYA HOPE Study Collaborative Group Received: 23 December 2010 /Accepted: 9 January 2011 /Published online: 28 January 2011 # The Author(s) 2011. This article is published with open access at Springerlink.com Abstract population-based sample of AYA survivors to examine Introduction Cancer is rare in adolescents and young adults issues of treatment and health outcomes. (AYA), but these patients have seen little improvement in Methods Individuals diagnosed in 2007–08 and age 15–39 survival in contrast to most other age groups. Furthermore, at the time of diagnosis with acute lymphocytic leukemia, participation in research by AYAs is typically low. We Hodgkin lymphoma, non-Hodgkin lymphoma, germ cell conducted a study to examine the feasibility of recruiting a cancer or sarcoma were identified by 7 Surveillance, The authors have no commercial associations, current and over the past 5 years, that might pose a conflict of interest. L. C. Harlan (*) A. W. Smith I. Kato National Cancer Institute, Karmanos Cancer Institute and Department of Pathology, Wayne State University School of Medicine, 6130 Executive Blvd, MSC 7344, Detroit, MI, USA Bethesda, MD 20892-7344, USA e-mail: lh50w@nih.gov R. D. Cress C. F. Lynch M. M. West Department of Public Health Sciences, Department of Epidemiology, University of Iowa, UC Davis School of Medicine, Iowa City, IA, USA Davis, CA, USA T. H. M. Keegan R. D. Cress Cancer Prevention Institute of California, Public Health Institute/California Cancer Registry, Fremont, CA, USA Sacramento, CA, USA T. H. M. Keegan S. M. Schwartz Department of Health Research and Policy, Public Health Sciences Division, Stanford University School of Medicine, Fred Hutchinson Cancer Research Center, Stanford, CA, USA Seattle, WA, USA A. S. Hamilton D. Deapen Department of Preventive Medicine, Keck School of Medicine, S. M. Stringer University of Southern California, Kelly Services, Scientific Resources, Los Angeles, CA, USA 6101 Executive Boulevard, Suite 392, Rockville, MD, USA X.-C. Wu Epidemiology Program/Louisiana Tumor Registry, School of A. L. Potosky Public Health, Louisiana State University Health Sciences Center, Lombardi Comprehensive Cancer Center, Georgetown University, New Orleans, LA, USA Washington, DC, USA 306 J Cancer Surviv (2011) 5:305–314 Epidemiology, and End-Results (SEER) cancer registries, health-related quality of life, unmet needs, and the impact of mailed surveys within 14 months after diagnosis and again cancer on psychosocial domains. The design and implemen- a year later, and had medical records reviewed. tation of the Adolescent and Young Adult Health Outcomes Results 525 (43%) of the eligible patients responded, 39% and Patient Experience (AYA HOPE) Study was motivated refused and 17% were lost to follow-up. Extensive efforts and guided by the findings of a systematic review of the were required for most potential respondents (87%). 76% science on AYA cancer patients [5]. of respondents completed the paper rather than online Here we report on the feasibility of 1) recruiting AYA survey version. In a multivariate model, age, cancer site, cancer survivors, 2) developing and fielding a patient education and months from diagnosis to the first mailing survey with a subsequent survey to examine changes over of the survey were not associated with participation, time and 3) obtaining patient (or guardian) consent to although males (p<0.01), Hispanics and non-Hispanic review medical records. We focus on the practical chal- blacks (p<0.001) were less likely to participate. 91% of lenges in recruiting and conducting research in AYA cancer survivors completing the initial survey completed the survivors identified through population-based registries subsequent survey. across the United States, and discuss potential strategies Discussion Despite the response rate, those who participated to increase recruitment of these survivors. adequately reflected the population of AYA cancer survivors. The study demonstrates that cancer registries are valuable foundations for conducting observational, longitudinal Methods population-based research on AYA cancer survivors. Implications for Cancer Survivors Achieving a reasonable Study setting and eligibility criteria response rate in this population is possible, but requires extensive resources. Patients were identified through seven population-based Surveillance, Epidemiology and End-Results (SEER) pro- . . Keywords Adolescent cancer Young adult cancer gram cancer registries: Detroit; Seattle/Puget Sound; Los . . . Survey Response rates Medical records Consent forms Angeles County, San Francisco/Oakland, Greater California (13 counties around Sacramento plus Orange County), and the states of Iowa and Louisiana. Introduction Conduct of this study required approval from 9 Institu- tional Review Boards (IRBs) (7 registries, 1 state (California) Although cancer is rare in adolescents and young adults and the NCI IRB), which took 7 months. Several registries (AYA), it is the leading cause of death in this age group experienced particularly long approval processes because the after unintentional injury, homicide and suicide [1]. inclusion of minors required changes to consent forms. No Survival following a cancer diagnosis in AYAs (defined as concerns were raised by the IRB about survey content, ages 15–39) has shown little improvement relative to other although there was concern about security of the online age groups [2]. In particular, cancer patients diagnosed survey that was resolved with a more complete explanation of between the ages of 25 and 35 have had no survival the security protocols. The online survey website included a improvement in more than 20 years [2]. Although scientific firewall, Virtual Private Network (VPN), an intrusion evidence to understand reasons for this phenomenon is detection system, and routine security checks of the computer limited, experts have speculated that the lack of improve- resources. Patients were given a website address, user name ment in survival may be due to a combination of factors, and password in the initial mailing. Patients accessing the such as the lack of available clinical trials hampering efforts online survey were required to create a new password upon to develop novel therapies,[3] lack of medical insurance, entering the system. This process made the survey accessible poor access to medical care for the initial diagnosis and for only to the person with the new password. Once the survey follow-up care, attitudes of invincibility, physicians’ low was submitted no further access was allowed. suspicion of cancer resulting in delayed diagnosis, and Eligible patients were diagnosed between July 1, 2007 location and specialty of treating physicians [4]. and October 31, 2008 with a first invasive, histologically To address several of these issues in AYA cancer patients, confirmed germ cell cancer, non-Hodgkin lymphoma the National Cancer Institute (NCI) in 2007, in partnership (NHL), Hodgkin lymphoma (HL), acute lymphocytic with the Lance Armstrong Foundation (LAF), initiated an leukemia, Ewing’s sarcoma, osteosarcoma, or rhabdomyo- observational cohort study to: 1) examine factors associated sarcoma (Appendix I), ages 15 through 39 years at with high quality cancer treatment in general community diagnosis, residents of the study area and able to read practice, including the use of clinical trials and treatment English. The goal was to obtain 530 surveys completed 6– protocols, and 2) assess patient-reported outcomes, such as 14 months from the date of diagnosis. A small sample of J Cancer Surviv (2011) 5:305–314 307 deceased patients who were otherwise eligible was included website directories (e.g., Google, Zabasearch), paid direc- in the medical record review. tory services (e.g., Accurint, Coles) and/or public records (e.g., Department of Motor Vehicles, Voter Registration). Data collection instruments A few study sites sought new addresses by exploring social networking websites. If a new address was not The survey (http://outcomes.cancer.gov/surveys/aya/), found, the staff attempted to contact the individuals by which took approximately 20 min to complete, was telephone. If the interviewer received an answering designed to be self-administered and address a number machine a message was left; after leaving 2 messages no of issues including the impact of cancer, health-related additional messages were left. Phone calls were made at quality of life, healthcare delivery, and reasons for not different times during the day, evening and weekends. If participating in clinical trials. Survey development included none of these efforts was successful, the patient was cognitive and usability testing with 24 AYA cancer survivor classified as lost to follow-up. volunteers. Patients who completed the survey were sent a thank Patients were asked to complete a healthcare provider you letter and $25.00 for the time spent completing it. form by listing the names and addresses of all healthcare An additional $25.00 was provided for time spent providers and facilities providing care. The information was completing the healthcare provider and medical record used to supplement registry information and to request release forms. medical records for data abstraction. Medical record retrieval Patient recruitment Medical records were collected from facilities listed on the Before patients were contacted, the patient’s physician was healthcare provider form. Information obtained included notified that his/her patient would be approached to type of healthcare facilities, physician subspecialties, tumor participate in the study. Physicians were asked to advise characteristics and staging, diagnostic procedures, partici- study staff if there were reasons the patient should not be pation in active clinical trials, therapy provided, and contacted. If the staff had not heard from the physician in comorbid conditions. 3 weeks, they contacted the patient. Some physicians and one state registry required active physician consent. Follow-up survey Eligible patients were mailed a cover letter, a brochure introducing the study, an overall study consent document We also conducted a follow-up survey, designed to be self- administered by paper or online, to examine changes in (where required by the registry’s IRB), healthcare provider form, a Health Insurance Portability and Accountability Act psychosocial outcomes and health-related quality of life (HIPAA) compliant consent for the release of medical 15–35 months following the initial diagnosis. Survey information, a paper survey, a web address for the online development (http://outcomes.cancer.gov/surveys/aya/), survey, a pre-addressed stamped return envelope for study tracing and follow-up was similar to the methods used for materials, and a LIVESTRONG bracelet. If the patient was the initial survey. Participants were provided with $50 for a minor at the time of contact, study materials were mailed time spent completing the survey. No additional medical to the parent/guardian and included a consent document record review was performed. requiring signatures from the parent/guardian and a signa- ture of assent from the minor. Analyses Follow-up procedures for non-respondents included a second mailing 3 weeks after the initial mailing, followed Descriptive univariate analyses were conducted to com- by telephone calls from trained interviewers beginning pare demographic and tumor characteristics of baseline 2 weeks after the second mailing. If needed, multiple survey respondents and non-respondents based on data follow-up calls were made to ask the patient to complete collected by each SEER registry. We used the patient’s the survey over the telephone and mail the medical record address at the time of cancer diagnosis to determine area release and the healthcare provider form. level educational attainment and median family income at If study materials were returned marked undeliverable, the census tract level [6]. A multiple logistic regression the staff employed a variety of tracing procedures to obtain model was used to assess the association between clinical and the correct address, including reviewing the cancer registry non-clinical factors and participation. Survey participants database for updated information, contacting directory were compared by survey mode, and follow-up survey assistance, obtaining information from healthcare providers, completion. A significance level of 0.05 (two tailed) was hospital(s) and family members. Study sites also used used for all analyses. 308 J Cancer Surviv (2011) 5:305–314 Results remaining 83 patients, neither a valid telephone number nor an address could be found. A total of 1,405 patients were identified as potentially eligible (Fig. 1). Among registries that required physician Generalizability of the enrolled sample consent, only 6 physicians refused to allow their patient to be contacted. Physicians identified 11 patients as We compared the characteristics of patients who participated deceased and 9 as ineligible. Another 70 patients became in the survey to non-responders (Table 1). In univariates ineligible because they exceeded their 14-month eligibil- analyses, non-respondents were not significantly different ity date prior to contact. After mailing the survey (n= from responders by age, census tract education, or census 1,309), we identified 16 deceased patients and 85 patients tract median family income. Response did vary by cancer ineligible because they did not speak English (56%), site (p<0.04) from 38% for acute lymphocytic leukemia and they denied having cancer (9%), or due to other reasons sarcoma to 51% for HL. Females were more likely to (34%). participate (p<0.0001). Non-Hispanic black (p<0.05) and Hispanic (p<0.001) patients were less likely to respond than Response rates non-Hispanic whites. Among non-respondents, non- Hispanics whites and non-Hispanic Blacks were more likely A total of 524 eligible patients completed the initial survey to refuse (p<0.001) while Hispanics (32%) were more than and 1 patient completed only the medical record release twice as likely to be lost to follow-up as non-Hispanic white (Fig. 1), yielding a response rate of 43%. Of patients sent (12%) or non-Hispanic black (14%) patients (data not the initial mailing, 87% required additional contact. shown). Of the refusals, 133 directly declined participation when In a multivariate logistic regression model, we investi- contacted and 341 patients never completed the surveys gated factors thought to be associated with response rate despite repeated contacts (Fig. 1). Of patients lost to follow- (Table 2). Females were more likely to answer the survey up, telephone numbers were found for 126, but several calls than were males (p <0.001). Compared to non-Hispanic to these numbers did not produce a mailing address. For the whites, Hispanics and non-Hispanic blacks were less likely Fig. 1 Patient number AYA HOPE PATIENT RECRUITMENT FLOWCHART flowsheet for the study Deceased Potential Eligible Patients Physician Consent N=11 N=1405 Refused N=6 Ineligible N=79 Deceased Refusals Potential Participants Mailed N=16 N=474 Survey N=1309 Ineligible Lost to Follow-up N=85 N=209 Total Eligible Surveys Completed N=524 Medical Record Consent only Lost Survey N=1 N=1 Paper Web Telephone N=397 N=115 N=12 Medical Record Medical Record Consent Refusals Consents N=26 N=499 Medical Records Obtained N=490 J Cancer Surviv (2011) 5:305–314 309 Table 1 Percent distribution of demographic characteristics for respondents and non-respondents ALL Germ cell HL NHL Sarcoma All cancer sites p-value R NR R NR R NR R NR R NR R NR (n=21) (n=35) (n=206) (n=304) (n=142) (n=136) (n=131) (n=167) (n=25) (n=41) (n=525) (n=683) Age (years) p=0.0788 15–19 57.1 42.9 7.8 12.5 12.7 20.6 9.9 9.6 32.0 56.1 12.8 17.6 20–24 14.3 28.6 21.4 18.4 16.9 22.1 12.2 14.4 24.0 17.1 17.7 18.6 25–29 14.3 11.4 27.2 24.0 31.7 19.1 19.1 18.6 20.0 9.8 25.5 20.2 30–34 0 8.6 24.3 24.0 21.1 18.4 25.2 24.0 12.0 12.2 22.1 21.4 35–39 14.3 8.6 19.4 21.1 17.6 19.9 33.6 33.5 12.0 4.9 21.9 22.3 SEER Race/ethnicity p=0.0026 Hispanic 33.3 31.4 22.3 33.2 12.0 11.8 17.6 21.6 16.0 29.3 18.5 25.8 Non-Hispanic White 47.6 51.4 70.4 53.6 62.7 65.4 59.5 44.3 40.0 41.5 63.2 52.9 Non-Hispanic Black 9.5 2.9 2.4 4.6 12.0 15.4 10.7 16.8 16.0 12.2 8.0 10.1 Other/Unknown 9.5 14.3 4.9 8.6 13.4 7.4 12.2 17.4 28.0 17.1 10.3 11.3 Sex p<0.0001 Male 47.6 65.7 90.3 90.8 39.4 56.6 49.6 60.5 64.0 68.3 63.4 73.9 Female 52.4 34.3 9.7 9.2 60.6 43.4 50.4 39.5 36.0 31.7 36.6 26.1 Census: Median% HS Education (age 25 or older) p=0.0945 Not Available 0 0 0.5 0 0 1.5 0 0 0 0 0.2 0.3 <75% 23.8 25.7 28.2 33.2 26.1 24.3 26.0 28.1 24.0 41.5 26.7 30.3 75–90% 52.4 45.7 37.4 43.4 41.5 41.9 35.1 38.3 52.0 29.3 39.2 41.1 90+% 23.8 28.6 34.0 23.4 32.4 32.4 38.9 33.5 24.0 29.3 33.9 28.3 Census: median family income p=0.1102 Not available 0 0 0.5 0 0 1.5 0 0 0 0 0.2 0.3 <$45,000 19.0 34.3 31.1 38.5 28.9 28.7 29.0 33.5 40.0 39.0 29.9 35.1 $45–$65,000 47.6 34.3 36.9 34.2 36.6 30.9 38.9 32.9 32.0 26.8 37.5 32.8 $65,000+ 33.3 31.4 31.6 27.3 34.5 39.0 32.1 33.5 28.0 34.1 32.4 31.8 R respondent; NR non-respondent; ALL acute lymphocytic leukemia; HL Hodgkin lymphoma; NHL non-Hodgkin lymphoma p-value=0.0372 to respond. Other factors in Table 2 were not significantly Medical records were obtained for 93% of all respondents. associated with participation. However, we were unable to obtain medical records for 2% of consenting participants because the records were Response by survey mode lost within the healthcare facility or the facility failed to provide them after multiple requests. Medical records Some patients chose to complete the survey online (22%) were obtained for a small sample (n=27) of deceased or over the telephone (2%) (Fig. 1 and Table 3). Compared patients primarily to assess any differences in the diagnosis to patients in other age or racial/ethnic groups, fewer 20– and treatment that might arise from excluding patients who 24 year-olds or non-Hispanic black patients and less than 10% had died. of sarcoma patients completed the survey online. Patients with an Associate Degree or higher or who lived in higher Follow-up survey socioeconomic status census tracts, defined by education and income, were more likely to complete the survey online. Of the 524 eligible participants who completed the initial survey, 465 completed the follow-up survey, 34 were non- Medical records responders, and 10 were lost to follow-up (Table 4). The participation rate in the follow-up survey was over 91% Among participants completing the survey, 5% refused to among survivors. Between the initial and subsequent sign a medical records release consent form (Fig. 1). survey (8–17 months), 3% (n=15) of the patients died. 310 J Cancer Surviv (2011) 5:305–314 Table 2 Multivariable analysis examining characteristics associated one recent health survey in adolescents ages 13–17 years with response vs. non-response to the AYA HOPE Study Baseline Survey yielded a 40% response rate [8]. These adolescents had been seen within the healthcare system in the previous year 95% Confidence Interval and were recruited by that system. A national study of childhood cancer survivors between 1994 and 2000 Variable OR Lower Upper p-value achieved a 62% completion rate with extensive resources [9]. However, the denominator for the completion rate Age at diagnosis (years) 0.06 included only those subjects who expressed an interest in 15–19 1.0 the study after being contacted by their treating institution. 20–24 1.3 0.9 2.0 As indicated by our follow-up survey, once individuals 25–29 1.8 1.2 2.7 agree to participate in a study, they are more likely to 30–34 1.5 0.97 2.2 complete subsequent surveys. The response rate in our 35–39 1.4 0.9 2.0 study might have been higher if, as in the NHL [7] study, SEER Racial/ethnicity 0.01 more time had elapsed since diagnosis. Some of the patients Non-Hispanic White 1.0 eligible for our study who refused may still have been Hispanic 0.6 0.5 0.9 under active treatment and may have agreed to participate Non-Hispanic Black 0.6 0.4 0.9 once therapy was completed. Other/Unknown 0.7 0.5 1.0 Males and non-Hispanic black and Hispanic patients SEER Sex <0.001 were significantly less likely to participate, similar to Male 1.0 differences reported for cancer clinical trials [10, 11]. Female 1.7 1.3 2.2 Significantly lower enrollment for minority compared to Cancer site 0.60 white patients and for men compared to women was Acute lymphocytic leukemia 1.0 reported for surgical trials [11]. This finding has been Germ cell cancer 1.1 0.6 2.0 shown not only in cancer, but in other diseases as well [12]. Hodgkin lymphoma 1.4 0.7 2.5 Although 37% of our respondents were racial/ethnic Non-Hodgkin lymphoma 1.1 0.6 2.0 minorities, it is likely to have been higher if translated Sarcoma 1.1 0.5 2.3 versions of the survey were available, as not speaking Census:% HS education 0.31 English was the primary reason for ineligibility among our (age 25 or older) population. <75% 1.0 A variety of methods were used to maximize responses 75–90% 0.9 0.7 1.2 to the mailing. We enclosed a LIVESTRONG bracelet that 90%+ 1.1 0.8 1.6 added interest and bulk to the packet. One registry initially Months from diagnosis to 1.0 0.9 1.01 0.14 used an overnight delivery service, but abandoned this initial mailing of survey within a month. Response rates were lower than with OR>1.0 indicates more likely to respond; OR<1.0 indicates less USPS. The overnight service did not leave packages if no likely to respond. one was home, had limited re-delivery, no forwarding addresses, and deliveries were limited to weekdays. While Discussion we used a variety of approaches to enhance our response rates, we did not send pre-notification letters or postcards. The goal of the AYA HOPE Study was to learn about Research has suggested this may not increase the response recruitment of AYAs, their cancer care and outcomes and to rate [12], but it would have identified incorrect addresses at determine the feasibility of collecting survey and detailed a lower cost than mailing the complete packet. medical record information on a representative population- Overall, 87% of the identified sample required at least based sample of AYA cancer survivors. Our response rate one additional mailing or telephone contact. An address or among eligible patients was 43%. This age group, partic- telephone number could not be found for 16% of patients ularly those ages 15–25 years, is mobile and difficult to identified as eligible, similar to the 15% reported by the follow due to educational and employment opportunities, childhood cancer survivors study [9]. Although we likely marriage and other personal life changes. had the correct phone number for some patients, no one Previous studies have shown that response rates of answered our calls. With the increased use of caller young adults are lower than for older adults [7–9]. A survey identification, patients may have been screening calls from of adult NHL survivors identified from a cancer registry unrecognized phone numbers. It is possible that calls had a 55% response rate, but patients were older and 2– originating from patients’ own medical institutions may 5 years post diagnosis [7]. Despite aggressive follow-up, have yielded better responses, although a response rate of J Cancer Surviv (2011) 5:305–314 311 Table 3 Mode of survey Paper Online Phone completion by patient characteristics, AYA HOPE N % N% N% study Age at diagnosis (years) 15–19 51 76.1% 14 20.9% 2 3.0% 20–24 75 80.6% 14 15.1% 4 4.3% 25–29 99 73.9% 31 23.1% 4 3.0% 30–34 84 73.0% 29 25.2% 2 1.7% 35–39 88 76.5% 27 23.5% 0 SEER Race/ethnicity Hispanic 75 78.1% 19 19.8% 2 2.1% Non-Hispanic White 252 75.9% 72 21.7% 8 2.4% Non-Hispanic Black 34 81.0% 6 14.3% 2 4.8% Other/Unknown 36 66.7% 18 33.3% 0 SEER Sex Male 240 72.3% 82 24.7% 10 3.0% Female 157 81.8% 33 17.2% 2 1.0% Cancer site ALL 18 85.7% 3 14.3% 0 NHL 95 72.5% 32 24.4% 4 3.1% HL 108 76.1% 32 22.5% 2 1.4% Germ cell 153 74.6% 46 22.4% 6 2.9% Sarcoma 23 92.0% 2 8.0% 0 Education Less than high school diploma 53 91.3% 4 6.9% 1 1.7% Completed high school 69 75.8% 16 17.6% 6 6.6% Some college/vocational/training 106 76.8% 30 21.7% 2 1.4% Associate Degree 38 71.7% 15 28.3% 0 College graduate 94 71.8% 34 26.0% 3 2.3% Post-graduate education 35 68.6% 16 31.4% 0 Missing 2 100.0% 0 0 Census:% HS Education (age 25 or older) Not Available 1 100.0% 0 0 <75% 109 78.4% 24 17.3% 6 4.3% 75–90% 166 80.6% 35 17.0% 5 2.4% 90%+ 121 68.0% 56 31.5% 1 0.6% Census: median family income Not Available 1 100.0% 0 0 <$45,000 127 81.4% 22 14.1% 7 4.5% $45,000–$65,000 149 75.6% 45 22.8% 3 1.5% $65,000+ 120 70.6% 48 28.2% 2 1.2% SEER Registry Northern California 81 65.9% 38 30.9% 4 3.3% Los Angeles 77 81.1% 17 17.9% 1 1.1% Seattle-Puget Sound 63 74.1% 22 25.9% 0 Louisiana 52 74.3% 12 17.1% 6 8.6% a Greater California 43 72.9% 16 27.1% 0 ALL acute lymphocytic leukemia; HL Hodgkin Metropolitan Detroit 44 84.6% 7 13.5% 1 1.9% lymphoma; NHL non-Hodgkin Iowa 37 92.5% 3 7.5% 0 lymphoma 312 J Cancer Surviv (2011) 5:305–314 Table 4 Follow-up survey Survey Completed Non-Response Deceased Lost to follow-up completion by patient characteristics, AYA HOPE N% N % N % N % Study Age at diagnosis (years) 15–19 63 94.0% 1 1.5% 0 3 4.5% 20–24 81 87.1% 8 8.6% 2 2.2% 2 2.2% 25–29 114 85.1% 13 9.7% 5 3.7% 2 1.5% 30–34 103 89.6% 6 5.2% 5 4.3% 2 1.7% 35–39 104 90.4% 7 6.1% 3 2.6% 1 0.9% SEER Race/ethnicity Hispanic 83 86.5% 6 6.2% 3 3.1% 5 5.2% Non-Hispanic White 298 89.8% 22 6.6% 8 2.4% 4 1.2% Non-Hispanic Black 35 83.3% 4 9.5% 2 4.8% 1 2.4% Other/Unknown 49 90.7% 3 5.6% 2 3.7% 0 SEER Sex Male 287 86.4% 26 7.8% 11 3.3% 9 2.7% Female 178 92.7% 9 4.7% 4 2.1% 1 0.5% Cancer site ALL 18 85.7% 1 4.8% 1 4.8% 1 4.8% NHL 113 86.3% 9 6.9% 7 5.3% 2 1.5% HL 130 91.5% 10 7.0% 1 0.7% 1 0.7% Germ cell 181 88.3% 15 7.3% 4 2.0% 6 2.9% Sarcoma 23 92.0% 0 2 8.0% 0 Education Less than high school diploma 51 87.9% 3 5.2% 3 5.2% 1 1.7% Completed high school 79 86.8% 8 8.8% 1 1.1% 3 3.3% Some college/vocational training 123 89.1% 8 5.8% 4 2.9% 3 2.2% Associate degree 47 88.7% 5 9.4% 1 1.9% 0 College graduate 119 90.8% 7 5.3% 4 3.1% 1 0.8% Post-graduate education 45 88.2% 3 5.9% 2 3.9% 1 2.0% Missing 1 50.0% 0 0 1 50.0% Census:% HS education (age 25 or older) Not Available 1 100.0% 0 0 0 <75% 121 87.1% 9 6.4% 5 3.6% 5 3.6% 75–90% 182 88.3% 14 6.8% 7 3.4% 3 1.5% 90%+ 161 90.4% 12 6.7% 3 1.7% 2 1.1% Census: median family income Not Available 1 100.0% 0 0 0 <$45,000 131 84.0% 13 8.3% 6 3.8% 7 4.5% $45,000–$65,000 177 89.8% 14 7.1% 5 2.5% 1 0.5% $65,000+ 156 91.8% 8 4.7% 4 2.4% 2 1.2% SEER Registry Northern California 108 87.8% 9 7.3% 2 1.6% 4 3.3% Los Angeles 81 85.3% 7 7.3% 4 4.2% 4 4.2% Seattle-Puget Sound 76 89.4% 6 7.1% 3 3.5% 0 Louisiana 60 85.7% 6 8.6% 4 5.7% 0 a Greater California 57 96.6% 0 1 1.7% 1 1.7% ALL acute lymphocytic leukemia; HL Hodgkin Metropolitan Detroit 49 94.2% 1 1.9% 1 1.9% 1 1.9% lymphoma; NHL non-Hodgkin Iowa 34 85.0% 6 15.0% 0 0 lymphoma J Cancer Surviv (2011) 5:305–314 313 only 40% was reported by a study conducted within dinal population-based research on younger, non-pediatric patients in a healthcare organization [8]. cancer survivors. Some survivors refused to participate in any part of our study because of the HIPAA medical release form. Open Access This article is distributed under the terms of the Evidence from a community-based, randomized, mixed- Creative Commons Attribution Noncommercial License which per- mits any noncommercial use, distribution, and reproduction in any mode survey (n=6,939) reported the inclusion of a medium, provided the original author(s) and source are credited. minimally burdensome version of HIPAA authorization form reduced survey response rates by up to 15% [13]. Simply requiring a signature reduces the response rate. Appendix I Nelson reported that response rates were 58% at locations requiring no advanced permission to contact the individual Sites and tumor morphologies (ICD-O-3 codes) included: compared to 27% for those requiring written permission from the individual [14]. non-Hodgkin lymphoma (9590–9596, 9670, 9671, Although some staff used social networking websites to 9673, 9675, 9678–9680, 9684, 9687, 9689–9691, find addresses, we did not use this approach to communicate 9695, 9698–9702, 9705, 9708, 9709, 9714–9719, with patients because registries did not request IRB approval 9727–9729); for this type of contact. We were concerned about the Hodgkin lymphoma (9650–9655, 9659, 9663–9665, confidentiality of a social network contact and identification 9667); of the correct person. However, in future studies, social germ cell cancer (9060–9091, 9100–9102) excluding networking websites may be a useful method of contact [15]. patients with intracranial and intraspinal tumors (C700- This may increase participation as email/web communication C729, C751-C753); is used extensively among young people. However, methods acute lymphocytic leukemia (9820, 9832–9837); to ensure patient confidentiality are required. Ewing’s sarcoma (9260, 9365), osteosarcoma (9180– The majority of patients in our study completed the 9187, 9192–9195) and rhabdomyosarcoma (8900–8921): paper rather than the online version. Future studies might Sarcomas arising in the central nervous system (C700- evaluate different approaches to increase participation and C729) were excluded. the use of online surveys. A web address included in the initial mailing that links to a well-designed website describing the study and reasons to participate, and includes References the survey as well as other survivorship information might entice AYAs to participate. However, it is possible that 1. Heron MP. Deaths: Leading causes for 2004. National vital AYAs simply prefer to complete the paper version of the statistics reports: vol 56 no 5. Hyattsville, MD: National Center survey. A study in Olmstead County found the inclusion of for Health Statistics, 2007. 2. Bleyer A, Budd T, Montello M. Adolescents and young adults an internet option decreased the response rate [16] while a with cancer: the scope of the problem and criticality of clinical study in Norway found no increase with the online option trials. Cancer. 2006;107(7 Suppl):1645–55. [17]. One possibility is that unless thrown away, a paper 3. Ferrari A, Montello M, Budd T, Bleyer A. The challenges of survey is a constant reminder, whereas a computer can be clinical trials for adolescent and young adults with cancer. Pediatr Blood Cancer. 2008;50:1101–04. turned off. 4. Loberiza FR, Cannon AJ, Weisenburger DD, Vose JM, Moehr MJ, Bast MA, et al. Survival disparities in patients with lymphoma according to place of residence and treatment provider: a Conclusion population-based study. J Clin Oncol. 2009;27:5376–82. 5. Albritton K, Caligiuri M, Anderson B, Nichols C, Ulman D, Adams H, et al. Closing the Gap: Research and Care Imperatives The AYA HOPE Study demonstrates that recruiting and for Adolescents and Young Adults with Cancer: Report of the following a diverse population of AYA survivors diagnosed Adolescent and Young Adult Oncology Progress Review Group. US DHHS, NIH, NCI, LIVESTRONG Young Adult Alliance, with different cancers and living throughout the US can be 108 pages [report on internet] available from: http://women. accomplished using population-based registries. However, cancer.gov/library/AYAO_PRG_Report_2006_FINAL.pdf. it is important for researchers to be realistic in their 6. http://www.census.gov/main/www/cen2000.html. expectations, recognizing that this population is mobile 7. Bellizzi KM, Miller MF, Arora NK, Rowland JH. Positive and negative life changes experienced by survivors of non-Hodgkin’s and difficult to contact. Achieving a high response rate is lymphoma. Ann Behav Med. 2007;34:188–99. challenging, requiring more extensive resources for follow- 8. Richards J, Wiese C, Katon W, Rockhill C, McCarty C, Grossman up. Despite these limitations and challenges, the AYA D, et al. Surveying Adolescents enrolled in a regional health care HOPE Study confirms that the use of cancer registries is a delivery organization: mail and phone follow-up—What works at valuable foundation for conducting observational, longitu- what cost? JABFM. 2010;23:534–41. 314 J Cancer Surviv (2011) 5:305–314 9. Robison LL, Mertens AC, Boice JD, Breslow NE, Donaldson SS, 17. Brogger J, Nystad W, Cappelen I, Bakke P. No increase in Green DM, et al. Study design and cohort characteristics of the response rates by adding a web response option to a postal childhood cancer survivor study: a multi-institutional collabora- population survey: A randomized trial. J Med Internet Res. 2007;9 tive project. Med Pediatr Oncol. 2002;38:229–39. (5):e40. Available from: http://www.jmir.org/2007/e40/. 10. Bleyer A, Budd T, Montello M. Older adolescents and young adults with cancer and clinical trials: Lack of participation and progress in North America. In: Bleyer A, Barr R, editors. Cancer in adolescent and young adults. Heidelburg: Springer Verlag; 2007. p. 71–81. AYA HOPE study collaborative group 11. Stewart JH, Bertoni AG, Staten JL, Levine EA, Gross CP. Participation in surgical oncology clinical trials: gender-, race/ethnicity-, and age- based disparities. Ann Surg Oncol. 2008;15:1818–9. California Cancer Registry/Public Health Institute (Sacramento, 12. Beebe TJ, Rey E, Ziegenfuss JY, Jenkins S, Lackore K, Talley CA): Rosemary Cress, DrPH (P.I.); Gretchen Agha; Mark Cruz NJ, et al. Shortening a survey and using alternative forms of Fred Hutchinson Cancer Research Center (Seattle, WA): Stephen prenotification: impact on response rate and quality. BMC Med M. Schwartz, Ph.D. (P.I.); Martha Shellenberger; Tiffany Janes Res Methodol 2010; 10:50. Available from: http://www. Karmanos Cancer Center (Detroit, MI): Ikuko Kato, Ph.D (P.I.); biomedcentral.com/1471-2288/10/50. Ann Bankowski; Marjorie Stock 13. Beebe TJ, Talley NJ, Camilleri M, Jankins SM, Anderson KJ, Louisiana State University (New Orleans, LA): Xiao-cheng Wu, Locke GR. The HIPAA authorization form and effects on survey M.D., MPH (P.I.); Vivien Chen, Ph.D.; Bradley Tompkins response rates, non-response bias, and data quality: a randomized Cancer Prevention Institute of California (Fremont, CA): Theresa community study. Med Care. 2007;45:959–65. Keegan, Ph.D, M.S. (P.I.); Laura Allen; Zinnia Loya; Karen Hussain 14. Nelson K, Elena Garcia R, Brown J, Mangione CM, Louis TA, University of Iowa (Iowa City, IA): Charles F. Lynch M.D., Ph.D. Keller E, et al. Do patient consent procedures affect participation (P.I.); Michele M. West, Ph.D.; Lori A. Odle, R.N. rates in health services research? Med Care. 2002;40:283–8. University of Southern California (Los Angeles, CA): Ann 15. Lenhart A, Purcell K, Smith A, Zickuhr K. Social media & mobile Hamilton, Ph.D (P.I.); Jennifer Zelaya; Mary Lo; Urduja Trinidad Internet use among teens and young adults. In Pew Internet & National Cancer Institute (Bethesda, MD): Linda C. Harlan, BSN, American Life Project. Pew Research Center: Washington, DC; 2010. MPH, Ph.D.; (Investigator) Ashley Wilder Smith, Ph.D, MPH (Co- 16. Ziegenfuss JU, Beebe TJ, Rey E, Schleck C, Locke GR, Talley investigator); Sonja M. Stringer, MPH; Gretchen Keel, BS,BA. NJ. Internet option in a mail survey: more harm than Good? Consultants: Arnold Potosky, PhD.; Keith Bellizzi, Ph.D.; Karen Epidemiology. 2010;21:585–86. Albritton, MD, Michael Link, MD; Brad Zebrack, Ph.D.,

Journal

Journal of Cancer SurvivorshipPubmed Central

Published: Jan 28, 2011

There are no references for this article.