Search

The statistical technique of multiple regression, commonly referred to as "multivariable regression," is often used in clinical research to quantify the relationships between multiple predictor variables and a single outcome variable of interest. The foundational theory underpinning multivariable regression assumes that all predictor variables are independent of one another. In other words, the effect of each independent variable is measured by its contribution to the regression equation while all other variables remain unchanged. In the presence of correlations between two or more variables, however, it is impossible to change one variable without a consequent change in the variable(s) it is linked to. This condition, known as "multicollinearity," can introduce errors into multivariable regression models by affecting estimates of the regression coefficients that quantify the relationship between individual predictor variables and the outcome variable. Errors that arise due to violations of the multicollinearity assumption are of special interest to radiation oncology researchers. Because of high levels of correlation among variables derived from points along individual organ dose-volume histogram (DVH) curves, as well as strong intercorrelations among dose-volume parameters in neighboring organs, dosimetric analyses are particularly subject to multicollinearity errors. For example, dose-volume parameters for the heart are strongly correlated not only with other points along the heart DVH curve but are likely also correlated with dose-volume parameters in neighboring organs such as the lung. In this paper, we describe the problem of multicollinearity in accessible terms and discuss examples of violations of the multicollinearity assumption within the radiation oncology literature. Finally, we provide recommendations regarding best practices for identifying and managing multicollinearity in complex data sets. [ABSTRACT FROM AUTHOR]

Purpose: The purpose of this study was to characterize retracted studies within the field of radiation oncology.Methods and Materials: Computerized searches were performed in Ovid MEDLINE, PubMed, Ovid EMBASE, and the Cochrane Library up to May 2017 looking for retracted studies using the terms "retraction note," "retracted note," "withdrawn" and "radiotherapy," and "radiation oncology." Additional studies were identified by hand-searching 10 discipline-specific journals. Two authors independently screened papers and then extracted author demographics, journal characteristics, and retraction-specific variables.Results: Of the 58 studies identified, the most common reasons for retraction were misconduct (43%), methodological error (21%), authorship issues (5%), unknown causes (5%), and journal (administrative) errors (3%). A total of 13 systematic reviews or protocols (22%) were withdrawn from the Cochrane Library for being out-of-date or redundant. All but one retracted study and retraction notice were available in portable document format. Of the 57 retrieved papers, 79% were identified as retracted via in-text notations or watermarks. Overall median time to retraction was 44 months (interquartile range, 11-98 months). However, 42 studies (72%) were still cited after retraction notices were published.Conclusions: A retracted study within the field of radiation oncology remains a relatively uncommon event. Although promising, our data suggest that the majority of these retracted articles continue to be cited as valid research. As such, there is still a need for clinicians to remain vigilant with their academic rigor and good clinical research practices. There is an urgent need for publication houses to foster universal publishing standards along with discipline-specific retraction guidelines. [ABSTRACT FROM AUTHOR]

Patient understanding of radiotherapy (RT) processes and data regarding optimal approaches to patient education (PE) within radiation oncology (RO) are limited. Our objective was to evaluate PE practices and preferences of interprofessional RO providers to inform recommendations for delivering inclusive, accessible, and patient-centered education. An anonymous 17-item online survey, approved by an ethics review board, was administered to all members of the Radiation Oncology Education Collaborative Study Group (ROECSG) between 10/5/22 to 11/23/22. Respondent demographics, provider practices/preferences, and institutional practices were collected. Qualitative items explored key strategies, challenges, and desired resources for PE. Descriptive statistics summarized survey responses. Fisher's exact test compared PE practices by provider role and PE timing. Thematic analysis was used for qualitative responses. A total of 123 ROECSG members, including RO attendings (64%), RO trainees (21%), medical physicists (7%), physician assistants/nurses (2%), and radiation therapists (2%), completed the survey (31% response rate). Most practiced in an academic setting (86%) in North America (82%). The most common PE resources used were custom created institution-specific (61%) and electronic health system generated materials (38%). PE was delivered primarily by one-on-one teaching (72%), paper handouts (69%), and organizational websites (21%) (e.g., RTanswers.org). Almost half (41%) reported that PE practices differed based on type of clinical encounter, for example paper handouts for in-person visits and multimedia for virtual visits. The majority (86%) stated that their institution has disease site-specific PE materials, with nearly all having breast cancer materials (91%). Only 58% reported access to non-English PE materials. RO attendings/trainees were more likely than other team members to deliver PE at consultation (98% vs 71%, p = 0.03). PE practices amongst radiation oncologists differed according to the timing along the RT care path (consultation vs simulation vs first fraction, respectively): one-on-one teaching: 89% vs 49% vs 56%, p<0.01 and paper handouts: 69% vs 28% vs 16%, p<0.01. Key PE strategies included incorporating multimedia resources, personalizing delivery, and repetition at multiple timepoints by the interprofessional team. Limited time, inadequate administrative support, and lack of customized resources were identified as challenges in PE delivery. Interprofessional RO providers engage in PE, with most utilizing institution-specific materials. PE practices differ according to the type of clinical encounter and timing in the RT care path. Increased adoption of multimedia materials and partnerships with patients to tailor PE resources based on language, learning styles, and cultural preferences are needed to foster high-quality, patient-centered PE delivery. [ABSTRACT FROM AUTHOR]

The objective of this study was to determine the impact of the COVID-19 pandemic on academic productivity in oncology as measured by conference abstracts, journal publications and individual authorship trends. Using a reference time frame of 2018 to 2022, we obtained data on the number of abstracts and articles submitted and published from a selection of oncology conferences and journals. To assess individual authorship patterns, we randomly selected 200 articles from 2018 (i.e., the 'index paper') and tracked publications over subsequent years for the first or last authors. Linear time-trend analyses and independent two-sample t-tests were used to assess changes in academic productivity over time, and univariable and multivariable linear regression were used to analyze individual factors predictive of publication rates, including gender, continent, specialty, MD vs. non-MD, and career status (early if within 5 years of training completion vs. late for all others). Data on submitted and published abstracts were available from 5 and 7 conferences, respectively. Both abstract categories demonstrated decreasing values over time but not statistically significant: conference submissions decreased from 15,308 in 2018 to 13,623 in 2022, (p = 0.11), and published abstracts decreased from 13,111 to 11,848, respectively (p = 0.16). Journal submissions were available from 6 journals and increased from 14,142 in 2018 to a peak of 20,241 in 2020 (2018 vs. 2020: p<0.001), and then declined to 15,650 in 2021. Journal publications from 10 journals showed no clear trends over time (p = 0.64). For the author-level analysis, of the 200 authors randomly selected, the majority were male (66.5%), from North America (55.5%), with an MD degree (80.9%, 131/162) and late career (86.6%, 129/149). Most common specialties included surgery (29%), radiation oncology (18.5%), epidemiology/public health (11%) and medical oncology (10%). For articles authored per year, there was no linear trend detected (p = 0.51), although mean number of publications per author peaked in 2020/2021 (papers/year for 2018-2022: 18.1, 18.1, 20.1, 21.2, 19.6). On univariable analysis, factors significantly associated with increasing publication rates were male gender, last author position on index paper, late career status, MDs, speciality of surgery or public/health epidemiology, and authors from Asia (all p<0.01). On multivariable analysis, factors remaining significantly predictive were late career status, MDs, specialty of surgeons or public health/epidemiologists, and authors from Asia (all p<0.01). Conference submissions and publications trended downward from 2018 to 2022. Journal submissions peaked overall in 2020, but did not translate to increases in journal publication rates. Disparities in publication trends were found, based on speciality and geographic regions, including a negative impact on early-career researchers. [ABSTRACT FROM AUTHOR]

Academic physicians typically do not receive a salary based on individual publications; however, publications are necessary for promotion, and the impact of publication productivity among ranks on salary is unknown. The objective of this was to evaluate salary changes associated with publication productivity among academic radiation oncologists. Radiation oncologist faculty-level academic productivity data were obtained, including the h-index, m-index, number of papers, and number of citations, based on prior work (Holliday et al, 2014). The AAMC and MGMA Provider Compensation Data from 2017 provided percentile-level salaries. The delta in salary was calculated for each delta in rank the publication metrics among ranks. The primary outcome was delta in salary per delta in h-index; additional calculations were performed for delta in m-index and delta in publications. A total of 986 radiation oncologists were included. For assistant professor, median salary was $367,000, average h-index 6.8, m-index 0.68, 15.7 publications. For associate professor, median salary was $452,000, average h-index 14, m-index 0.87, 41.8 publications. For full professor, median salary was $520,000, average h-index 31.3, m-index 1.33, 118.7 publications. For chair (not mutually exclusive from full professor), median salary was $720,000, average h-index 34.8, m-index 1.36, 146.8 publications. The delta in salary per delta in rank and the associated changes in h-index, m-index, and publications are shown in Table 1. The average change in salary from assistant to associate professor is $11,805.56 / h-index, $447,368.42 / m-index, and $3,256.70 / publication; for associate to full it is $3,930.64 / h-index, $147,826.09 / m-index, and $884.27 / publication; and for full to chair it is $57,142.86 / h-index, and $7,117.44 / publication. This work provides the average change in salary among academic ranks based on changes in research publication productivity. Our present analysis is unable to assess causality of this association, and many unaccounted confounders may affect this relationship. Further work in this area may include evaluation of demographic factors that have demonstrated rank and salary disparities in radiation oncology, such as gender. [ABSTRACT FROM AUTHOR]

The discovery of X rays in 1895 captivated society like no other scientific advance. Radiation instantly became the subject not only of numerous scientific papers but also of circus bazaars, poetry, fiction, costume design, comics, and marketing for household items. Its spread was "viral." What is not well known, however, is its incorporation into visual art, despite the long tradition of medicine and surgery as a subject in art. Using several contemporary search methods, we identified 5 examples of paintings or sculpture that thematically feature radiation therapy. All were by artists with exhibited careers in art: Georges Chicotot, Marcel Duchamp, David Alfaro Siqueiros, Robert Pope, and Cookie Kerxton. Each artist portrays radiation differently, ranging from traditional healer, to mysterious danger, to futuristic propaganda, to the emotional challenges of undergoing cancer therapy. This range captures the complex role of radiation as both a therapy and a hazard. Whereas some of these artists are now world famous, none of these artworks are as well known as their surgical counterparts. The penetration of radiation into popular culture was rapid and pervasive; yet, its role as a thematic subject in art never fully caught on, perhaps because of a lack of understanding of the technology, radiation's intangibility, or even a suppressive effect of society's ambivalent relationship with it. These 5 artists have established a rich foundation upon which pop culture and art can further develop with time to reflect the extraordinary progress of modern radiation therapy. [ABSTRACT FROM AUTHOR]

Radiation oncology has always been deeply rooted in modeling, from the early days of isoeffect curves to the contemporary Quantitative Analysis of Normal Tissue Effects in the Clinic (QUANTEC) initiative. In recent years, medical modeling for both prognostic and therapeutic purposes has exploded thanks to increasing availability of electronic data and genomics. One promising direction that medical modeling is moving toward is adopting the same machine learning methods used by companies such as Google and Facebook to combat disease. Broadly defined, machine learning is a branch of computer science that deals with making predictions from complex data through statistical models. These methods serve to uncover patterns in data and are actively used in areas such as speech recognition, handwriting recognition, face recognition, "spam" filtering (junk email), and targeted advertising. Although multiple radiation oncology research groups have shown the value of applied machine learning (ML), clinical adoption has been slow due to the high barrier to understanding these complex models by clinicians. Here, we present a review of the use of ML to predict radiation therapy outcomes from the clinician's point of view with the hope that it lowers the "barrier to entry" for those without formal training in ML. We begin by describing 7 principles that one should consider when evaluating (or creating) an ML model in radiation oncology. We next introduce 3 popular ML methods--logistic regression (LR), support vector machine (SVM), and artificial neural network (ANN)--and critique 3 seminal papers in the context of these principles. Although current studies are in exploratory stages, the overall methodology has progressively matured, and the field is ready for larger-scale further investigation. [ABSTRACT FROM AUTHOR]

Purpose This paper reports long-term results of RTOG 9903, to determine whether the addition of erythropoietin (EPO) would improve the outcomes of radiation therapy (RT) in mildly to moderately anemic patients with head and neck squamous cell carcinoma (HNSCCa). Methods and Materials The trial included HNSCCa patients treated with definitive RT. Patients with stage III or IV disease received concomitant chemoradiation therapy or accelerated fractionation. Pretreatment hemoglobin levels were required to be between 9.0 and 13.5 g/dL (12.5 g/dL for females). EPO, 40,000 U, was administered weekly starting 7 to 10 days before RT was initiated in the RT + EPO arm. Results A total of 141 of 148 enrolled patients were evaluable. The baseline median hemoglobin level was 12.1 g/dL. In the RT + EPO arm, the mean hemoglobin level at 4 weeks increased by 1.66 g/dL, whereas it decreased by 0.24 g/dL in the RT arm. With a median follow-up of 7.95 years (range: 1.66-10.08 years) for surviving patients and 3.33 years for all patients (range: 0.03-10.08 years), the 5-year estimate of local-regional failure was 46.2% versus 39.4% ( P =.42), local-regional progression-free survival was 31.5% versus 37.6% ( P =.20), and overall survival was 36.9% versus 38.2% ( P =.54) for the RT + EPO and RT arms, respectively. Late toxicity was not different between the 2 arms. Conclusions This long-term analysis confirmed that despite the ability of EPO to raise hemoglobin levels in anemic patients with HNSCCa, it did not improve outcomes when added to RT. The possibility of a detrimental effect of EPO could not be ruled out. [ABSTRACT FROM AUTHOR]

Purpose: To determine the characteristics, needs, and concerns of the current radiation oncology workforce, evaluate best practices and opportunities for improving quality and safety, and assess what we can predict about the future workforce. Methods and Materials: An online survey was distributed to 35,204 respondents from all segments of the radiation oncology workforce, including radiation oncologists, residents, medical dosimetrists, radiation therapists, medical physicists, nurse practitioners, nurses, physician assistants, and practice managers/administrators. The survey was disseminated by the American Society for Radiation Oncology (ASTRO) together with specialty societies representing other workforce segments. An overview of the methods and global results is presented in this paper. Results: A total of 6765 completed surveys were received, a response rate of 19%, and the final analysis included 5257 respondents. Three-quarters of the radiation oncologists, residents, and physicists who responded were male, in contrast to the other segments in which two-thirds or more were female. The majority of respondents (58%) indicated they were hospital-based, whereas 40% practiced in a free-standing/satellite clinic and 2% in another setting. Among the practices represented in the survey, 21.5% were academic, 25.2% were hospital, and 53.3% were private. A perceived oversupply of professionals relative to demand was reported by the physicist, dosimetrist, and radiation therapist segments. An undersupply was perceived by physician's assistants, nurse practitioners, and nurses. The supply of radiation oncologists and residents was considered balanced. Conclusions: This survey was unique as it attempted to comprehensively assess the radiation oncology workforce by directly surveying each segment. The results suggest there is potential to improve the diversity of the workforce and optimize the supply of the workforce segments. The survey also provides a benchmark for future studies, as many changes in the healthcare field exert pressure on the workforce. [Copyright &y& Elsevier]

Purpose: Locoregional control rates, late normal tissue sequelae, and functional outcome scores have not been different for tonsillar fossa and/or soft palate tumors treated by either brachytherapy (BT) or surgery in an organ function preservation protocol. For additional prioritizing in clinical decision-making, we focused on a comparison of the full hospital costs of the different treatment options.Methods and materials: Between 1986 and 2001, tonsillar fossa and/or soft palate tumors were treated by external beam radiotherapy (EBRT) to the primary tumor and neck, followed by fractionated BT to the primary. Neck dissection (ND) was performed for node-positive disease (BT group; 104 patients). If BT was not feasible, resection combined with postoperative EBRT was executed (surgery group; 86 patients). Locoregional control, disease-free survival, and overall survival were calculated according to the Kaplan-Meier method. The performance status scales, late side effects, and degree of xerostomia have been previously reported. This paper focused on the hospital and follow-up costs for the treatment groups EBRT and BT with or without ND compared with surgery followed by postoperative RT (PORT). Finally, these costs were also computed for future treatment strategies (e.g., better sparing of normal tissues by intensity-modulated RT [IMRT]).Results: Locoregional control, disease-free survival, and overall survival rate at 5 years for patients treated with EBRT and BT with or without ND vs. surgery plus PORT was 80% vs. 78%, 58% vs. 55%, and 67% vs. 57%, respectively. The major late side effect was xerostomia. Dry mouth syndrome affected the BT group and surgery group equally. The total costs for all treatment groups were €14,262 (BT group), €16,628 (BT plus ND group), €18,782 (surgery plus PORT group), €14,532 (IMRT group), and €16,897 (IMRT plus ND group).Conclusion: Excellent locoregional tumor control was observed with either modality, with no statistically significant differences in the incidence of the most noted side effect xerostomia. The total costs for BT were less than for surgery: €16,628 ($19,452) for EBRT plus BT plus ND vs. €18,782 ($22,074) for surgery plus PORT. To reduce the morbidity of xerostomia, we propose further optimizing our organ function preservation protocol by implementing IMRT as a more conformal, tissue-sparing, RT technique. This is of particular interest because the costs of IMRT plus ND (€16,897; $19,767) were not very different from those for BT plus ND (€16,628; $19,452) and were far less than the costs for surgery. [Copyright &y& Elsevier]

Head and neck cancer is the fifth most common cancer in the United States, with an overall survival rate of approximately 40–50%. In an effort to improve patient outcomes, research efforts designed to maximize benefit and reduce toxicities of therapy are in progress. Basic research in cancer biology has accelerated this endeavor and provided preclinical data and technology to support clinically relevant advances in early detection, prognostic and predictive biomarkers. Recent completion of the Human Genome Project has promoted the rapid development of novel “omics” technologies that allow more broad based study from a systems biology perspective. However, clinically relevant application of resultant gene signatures to clinical trials within cooperative groups has advanced slowly. In light of the large numbers of variables intrinsic to biomarker studies, validation of preliminary data for clinical implementation presents a significant challenge and may only be realized with large trials that involve significant patient numbers. The Radiation Therapy Oncology Group (RTOG) Head and Neck Cancer Translational Research Program recognizes this problem and brings together three unique features to facilitate this research: (1) availability of large numbers of clinical specimens from homogeneously treated patients through multi-institutional clinical trials; (2) a team of physicians, scientists, and staff focused on patient-oriented head-and-neck cancer research with the common goal of improving cancer care; and (3) a funding mechanism through the RTOG Seed Grant Program. In this position paper we outline strategic plans to further promote translational research within the framework of the RTOG. [Copyright &y& Elsevier]

Purpose: To determine and analyze the dosimetric consequences of current portal imaging practices for pediatric patients, and make specific recommendations for reducing exposure from portal imaging procedures. Methods and Materials: A survey was sent to approximately 250 Children's Oncology Group (COG) member institutions asking a series of questions about their portal imaging practices. Three case studies are presented with dosimetric analysis to illustrate the magnitude of unintended dose received by nontarget tissues using the most common techniques from the survey. Results: The vast majority of centers use double-exposure portal image techniques with a variety of open field margins. Only 17% of portal images were obtained during treatment, and for other imaging methods, few centers subtract monitor units from the treatment delivery. The number of monitor units used was nearly the same regardless of imager type, including electronic portal imaging devices. Eighty-six percent imaged all fields the first week and 17% imaged all fields every week. An additional 1,112 cm3 of nontarget tissue received 1 Gy in one of the example cases. Eight new recommendations are made, which will lower nontarget radiation doses with minimal impact on treatment verification accuracy. Conclusion: Based on the survey, changes can be made in portal imaging practices that will lower nontarget doses. It is anticipated that treatment verification accuracy will be minimally affected. Specific recommendations made to decrease the imaging dose and help lower the rate of radiation-induced secondary cancers in children are proposed for inclusion in future COG protocols using radiation therapy. [ABSTRACT FROM AUTHOR]