Generating Real-World Tumor Burden Endpoints from Electronic Health Record Data: Comparison of RECIST, Radiology-Anchored, and Clinician-Anchored Approaches for Abstracting Real-World Progression in Non-Small Cell Lung Cancer.

INTRODUCTION: Real-world evidence derived from electronic health records (EHRs) is increasingly recognized as a supplement to evidence generated from traditional clinical trials. In oncology, tumor-based Response Evaluation Criteria in Solid Tumors (RECIST) endpoints are standard clinical trial metrics. The best approach for collecting similar endpoints from EHRs remains unknown. We evaluated the feasibility of a RECIST-based methodology to assess EHR-derived real-world progression (rwP) and explored non-RECIST-based approaches. METHODS: In this retrospective study, cohorts were randomly selected from Flatiron Health's database of de-identified patient-level EHR data in advanced non-small cell lung cancer. A RECIST-based approach tested for feasibility (N = 26). Three non-RECIST approaches were tested for feasibility, reliability, and validity (N = 200): (1) radiology-anchored, (2) clinician-anchored, and (3) combined. Qualitative and quantitative methods were used. RESULTS: A RECIST-based approach was not feasible: cancer progression could be ascertained for 23% (6/26 patients). Radiology- and clinician-anchored approaches identified at least one rwP event for 87% (173/200 patients). rwP dates matched 90% of the time. In 72% of patients (124/173), the first clinician-anchored rwP event was accompanied by a downstream event (e.g., treatment change); the association was slightly lower for the radiology-anchored approach (67%; 121/180). Median overall survival (OS) was 17 months [95% confidence interval (CI) 14, 19]. Median real-world progression-free survival (rwPFS) was 5.5 months (95% CI 4.6, 6.3) and 4.9 months (95% CI 4.2, 5.6) for clinician-anchored and radiology-anchored approaches, respectively. Correlations between rwPFS and OS were similar across approaches (Spearman's rho 0.65-0.66). Abstractors preferred the clinician-anchored approach as it provided more comprehensive context. CONCLUSIONS: RECIST cannot adequately assess cancer progression in EHR-derived data because of missing data and lack of clarity in radiology reports. We found a clinician-anchored approach supported by radiology report data to be the optimal, and most practical, method for characterizing tumor-based endpoints from EHR-sourced data. FUNDING: Flatiron Health Inc., which is an independent subsidiary of the Roche group.

Examination of computer task exposures in radiologists: a work systems approach.

Radiologists are intensive computer users as they review and interpret radiological examinations using the Picture Archiving and Communication Systems (PACS). Since their computer tasks require the prolonged use of pointing devices, a high prevalence of Musculoskeletal Disorders (MSDs) is reported. The first phase of this study involved conducting a Cognitive Work Analysis in conjunction with a Participatory Ergonomics approach to perform a total work system analysis. We also conducted an ergonomic survey as well as collected computer use data, specifically for the mouse and keyboard. The goal of the study was to reduce the physical exposures for radiologists. This paper presents Phase I results describing the analyses and redesign process of the radiologists tasks, training design, computer use, and selected survey results.

Daily self-reports resulted in information bias when assessing exposure duration to computer use.

BACKGROUND: Self-reported exposure duration to computer use is widely used in exposure assessment, and this study examined the associated information bias in a repeated measures setting. METHODS: For 3 weeks, 30 undergraduate students reported daily cumulative computer-use duration and musculoskeletal symptoms at four random times per day. Usage-monitor software installed onto participant's personal computers provided the reference measure. We compared daily self-reported and software-recorded duration, and modeled the effect of musculoskeletal symptoms on observed differences. RESULTS: The relationships between daily self-reported and software-recorded computer-use duration varied greatly across subject with Spearman's correlations ranging from -0.22 to 0.8. Self-reports generally overestimated computer use when software-recorded durations were less than 3.6 hr, and underestimated when above 3.6 hr. Experiencing symptoms was related to a 0.15-hr increase in self-reported duration after controlling for software-recorded duration. CONCLUSIONS: Daily self-reported computer-use duration had a weak-to-moderate correlation with software-recorded duration, and their relationship changed slightly with musculoskeletal symptoms. Self-reports resulted in both non-differential and differential information bias.

Computer use patterns associated with upper extremity musculoskeletal symptoms.

INTRODUCTION: Over half of surveyed college students are experiencing pain they are attributing to computer use. The study objective was to evaluate the effect of computing patterns on upper extremity musculoskeletal symptoms. METHODS: Symptom experiences and computing/break patterns were reported several times daily over three weeks for 30 undergraduate students over a semester. Two-level logistic regression models described the daily association between each computing pattern and both any and moderate or greater symptom experienced, adjusting for covariates. RESULTS: The associations between most computing/break patterns and experiencing any symptoms were positive: total hours of computer use adjOR = 1.1 (90% CI 1.1-1.2), 1-2 breaks versus none adjOR = 1.3 (90% CI 0.9-1.9), 3-6 breaks versus none adjOR = 1.5 (90% CI 1.1-2.2), >15 min break versus none adjOR = 1.6 (90% CI 1.1-2.2), and number of stretch breaks adjOR = 1.3 (90% CI 1.1-1.5). However, breaks for less than 15 min were negatively associated with experiencing any symptoms: adjOR = 0.6 (90% CI 0.5-0.9). The associations between most computing/break patterns and experiencing moderate or greater symptoms were positive: total hours of computer use OR = 1.1 (90% CI 1.1-1.2), 1-2 breaks and 5-6 breaks versus none OR = 1.8 (90% CI 1.1-2.9), 7-8 breaks versus none OR = 2.0 (1.0-4.2), >15 min break versus none 1.8 (1.1-3.1), and number of stretch breaks OR = 1.3 (1.0-1.5). CONCLUSION: Computing/break patterns were consistently associated with experiencing symptoms. Our findings suggest evaluating breaks with computing duration (computing patterns) is more informative than assessing computing duration alone and can be used to better design ergonomic training programs for student populations that incorporate break times.

A multi-method study evaluating computing-related risk factors among college students.

PURPOSE: To characterize undergraduate computer use using different data collection methods, emphasizing computing-related postures, use patterns and upper extremity musculoskeletal symptoms. SUBJECTS AND METHODS: In Spring, 2004, undergraduate students from a single dormitory at a private university agreed to complete a College Computing & Health Survey. For three separate data collection periods each lasting a week, we observed postures during computer once per period and continuously measured computer input device usage. During these three periods, students self-reported computer usage and symptoms 3-5 times daily. RESULTS: Thirty students participated and all completed the study. Eighty-six percent reported ever experiencing symptoms after computer work. There were no time-related trends across data collection periods for posture, symptoms, and computing activities and patterns. Typed work and communicating (when compared with playing games) were usually the predominant computing activities throughout the semester. There was significantly greater self-reported computer use than that directly measured (p<0.05). CONCLUSION: This is the first study that utilized several methods of exposure assessment to describe computing postures, use patterns and upper extremity musculoskeletal symptoms among a college student cohort. Epidemiological studies need to explore time-related changes such as time of day, weekday, and days into the semester to further understand symptoms, posture, and computer use changes.

Daily computer usage correlated with undergraduate students' musculoskeletal symptoms.

BACKGROUND: A pilot prospective study was performed to examine the relationships between daily computer usage time and musculoskeletal symptoms on undergraduate students. METHODS: For three separate 1-week study periods distributed over a semester, 27 students reported body part-specific musculoskeletal symptoms three to five times daily. Daily computer usage time for the 24-hr period preceding each symptom report was calculated from computer input device activities measured directly by software loaded on each participant's primary computer. General Estimating Equation models tested the relationships between daily computer usage and symptom reporting. RESULTS: Daily computer usage longer than 3 hr was significantly associated with an odds ratio 1.50 (1.01-2.25) of reporting symptoms. Odds of reporting symptoms also increased with quartiles of daily exposure. CONCLUSIONS: These data suggest a potential dose-response relationship between daily computer usage time and musculoskeletal symptoms.