Unidimensional Measurement Research Articles

Rewards of bridging the divide between measurement and clinical theory: Demonstration of a bifactor model for the Brief Symptom Inventory.

There is growing evidence that psychiatric disorders maintain hierarchical associations where general and domain-specific factors play prominent roles (see D. Watson, 2005). Standard, unidimensional measurement models can fail to capture the meaningful nuances of such complex latent variable structures. The present study examined the ability of the multidimensional item response theory bifactor model (see R. D. Gibbons & D. R. Hedeker, 1992) to improve construct validity by serving as a bridge between measurement and clinical theories. Archival data consisting of 688 outpatients' psychiatric diagnoses and item-level responses to the Brief Symptom Inventory (BSI; L. R. Derogatis, 1993) were extracted from files at a university mental health clinic. The bifactor model demonstrated superior fit for the internal structure of the BSI and improved overall diagnostic accuracy in the sample (73%) compared with unidimensional (61%) and oblique simple structure (65%) models. Consistent with clinical theory, multiple sources of item variance were drawn from individual test items. Test developers and clinical researchers are encouraged to consider model-based measurement in the assessment of psychiatric distress.

Latent variable mixture models: a promising approach for the validation of patient reported outcomes

A fundamental assumption of patient-reported outcomes (PRO) measurement is that all individuals interpret questions about their health status in a consistent manner, such that a measurement model can be constructed that is equivalently applicable to all people in the target population. The related assumption of sample homogeneity has been assessed in various ways, including the many approaches to differential item functioning analysis. This expository paper describes the use of latent variable mixture modeling (LVMM), in conjunction with item response theory (IRT), to examine: (a) whether a sample is homogeneous with respect to a unidimensional measurement model, (b) implications of sample heterogeneity with respect to model-predicted scores (theta), and (c) sources of sample heterogeneity. An example is provided using the 10 items of the Short-Form Health Status (SF-36(®)) physical functioning subscale with data from the Canadian Community Health Survey (2003) (N = 7,030 adults in Manitoba). The sample was not homogeneous with respect to a unidimensional measurement structure. Specification of three latent classes, to account for sample heterogeneity, resulted in significantly improved model fit. The latent classes were partially explained by demographic and health-related variables. The illustrative analyses demonstrate the value of LVMM in revealing the potential implications of sample heterogeneity in the measurement of PROs.

Aspects et évaluation post-thérapeutiques des lésions du foie après traitement non chirurgical

The main non-surgical treatments for liver lesions include chemotherapy, targeted treatments, chemoembolization and radiofrequency ablation. The post-treatment imaging features are variable and depend on the initial appearance of the lesion, the type of treatment and the imaging modality. Evaluation of tumour response to treatment is important. RECIST criteria based on unidimensional lesion measurements may not always be appropriate. Other evaluation criteria (Choi for GIST, EASL for HCC or Chun criteria.) may be more relevant.

Understandings and misunderstandings of multidimensional poverty measurement

Multidimensional measures provide an alternative lens through which poverty may be viewed and understood. In recent work we have attempted to offer a practical approach to identifying the poor and measuring aggregate poverty (Alkire and Foster, J Public Econ, 2011). As this is quite a departure from traditional unidimensional and multidimensional poverty measurement—particularly with respect to the identification step—further elaboration may be warranted. In this paper we elucidate the strengths, limitations, and misunderstandings of multidimensional poverty measurement in order to clarify the debate and catalyse further research. We begin with general definitions of unidimensional and multidimensional methodologies for measuring poverty. We provide an intuitive description of our measurement approach, including a ‘dual cutoff’ identification step that views poverty as the state of being multiply deprived, and an aggregation step based on the traditional FGT measures. We briefly discuss five characteristics of our methodology that are easily overlooked or mistaken and conclude with some brief remarks on the way forward.

A scientific approach to optimal treatment of cruciate ligament injuries

A scientific approach to optimal treatment of cruciate ligament injuries

Metrics for measuring tumor burden: A comparison of unidimensional and volumetric measurements.

e13596 Background: Response Evaluation Criteria In Solid Tumors (RECIST) is widely used to evaluate the efficacy of anticancer therapies. However, relying upon a unidimensional length to approximate change in tumor burden may be biased. The purpose of this study was to compare tumor responses/changes assessed by volumetric and RECIST unidimensional measurements. Methods: CT scans of chest, abdomen and pelvis were acquired at 5mm slice thickness as per study protocol, from 49 patients enrolled in early phase oncology trials. Baseline and the first follow-up scans (~6 weeks) were analyzed. Tumor greatest diameter was manually measured by a radiologist and tumor volume by semi-automated computer algorithms. Results: 144 individual sites of metastatic disease were analysed, including from liver, lung and lymph nodes. As expected, the magnitude of tumor burden change was larger for volume than unidimensional measurements: 32/49 (65%) patients had an increase in tumor volume of more than 30% compared to 6/49 (12%) for unidimensional measurements. Based upon RECIST and mathematical equivalence (assuming spherical tumors) of partial response (PR) and progressive disease (PD) for volume, 9/49 (18%) patients were discordant. Out of these 9 cases, 6/9 (67%) were classified as stable disease (SD) using unidimensional measurements and PD using volume, 2/9 (22%) were classified as PD for unidimensional measurements and SD using volume, and 1/9 (11%) was classified as a PR using unidimensional measurements and SD using volume. Comparing tumor burden changes from baseline with 95% limits of agreement (derived from repeat scan reads; ECR 2011) for unidimensional measurements (-12.4, 20.4%) and volumetric measurements (-9.6, 18.0%), 20/49 (41%) patients were discordant. Out of these 20 cases, 19/20 (95%) were classified as having ‘no change’ using unidimensional and ‘change’ using volume (3/20 [15%], decrease; 16/20 [80%], increase). 1/20 (5%) patient was classified as having ‘no change’ using volume and ‘change’ (a decrease) using unidimensional measurements. Conclusions: There was some evidence to suggest that volume is more sensitive than unidimensional measurements in identifying tumor changes for individual patients.

The Role of Syllables in Anagram Solution: A Rasch Analysis

ABSTRACT Anagrams are frequently used by experimental psychologists interested in how the mental lexicon is organized. Until very recently, research has overlooked the importance of syllable structure in solving anagrams and assumed that solution difficulty was mainly due to frequency factors (e.g., bigram statistics). The present study uses Rasch analysis to demonstrate that the number of syllables is a very important factor influencing anagram solution difficulty for both good and poor problem solvers, with polysyllabic words being harder to solve. Furthermore, it suggests that syllable frequency may have an impact on solution times for polysyllabic words, with more frequent syllables being more difficult to solve. The study illustrates the advantages of Rasch analysis for reliable and unidimensional measurement of item difficulty.

Longitudinal measurement invariance of the Multiple Sclerosis Walking Scale-12

Objective One primary assumption underlying the unambiguous interpretation of change in Multiple Sclerosis Walking Scale-12 (MSWS-12) scores over time is longitudinal measurement invariance (i.e., Is the same construct being measured over time?). Such an assumption was tested in the present study over periods of 6 and 12 months in persons with relapsing–remitting multiple sclerosis (RRMS). Method Participants completed a battery of questionnaires that included the MSWS-12 at baseline ( n = 269) and 6-months ( n = 260) and 12-months ( n = 252) follow-up. The data were analyzed using confirmatory factor analysis and a series of nested model comparisons in M plus 3.0. Results The results indicated that the unidimensional measurement model and all of its parameters (e.g., factor loadings and item intercepts) were invariant over periods of 6 and 12 months. Conclusion We provide novel evidence that supports the unambiguous interpretation of scores from the MSWS-12 as a measure of change in walking impairment over time in a sample of persons with MS.

Accuracy of unidimensional and volumetric ultrasound measurements in predicting good pathological response to neoadjuvant chemotherapy in breast cancer patients

Pathological complete response (pCR) is an important predictor of long-term survival in patients with breast cancer receiving neoadjuvant chemotherapy (NAC). At present, the accuracy of traditional radiological assessments during treatment in predicting pCR is poor. Unidimensional and 3D volumetric ultrasound measurements prior to, after 4 cycles (mid-treatment), and at the end of 8 cycles (end-treatment) of chemotherapy were available from a subset of 55 patients enrolled in Neo-tAnGo, a National Cancer Research Network (NCRN) UK neoadjuvant chemotherapy breast cancer trial. Proportional changes in longest diameter (LD) and volume as well as absolute residual size thresholds were examined for their ability to predict pCR or pCR plus minimal residual disease (pCR/MRD). Sensitivity, specificity, positive (PPV) and negative predictive values (NPV) and likelihood ratios (LRs) were calculated. Receiver-operator characteristic (ROC) curves and logistic regression models were also constructed. At mid-treatment, neither complete radiological response, nor proportional LD or volume changes were found predictive of final pCR. A small residual tumour volume (≤ 1 cm³ vs. > 1 cm³) at mid-treatment, however, was associated with pCR/MRD (P = 0.014). Sensitivity, specificity, PPV, NPV, LR+ and LR- values were 61%, 77%, 61%, 77%, 2.62 and 0.51, respectively. The area under the ROC curve was 0.689 (P = 0.03). Volume ≤ 1 cm³ at mid-treatment was found significant in a logistic regression (OR: 0.194, P = 0.011). At end-treatment, no ultrasound measurements were found predictive of pCR or pCR/MRD. In conclusion, proportional tumour size changes (the basis of the RECIST criteria) were not found predictive of good pathological response, although residual volume ≤ 1 cm³ at mid-treatment was found to be predictive of pCR/MRD. However, multiple volume and LD thresholds were examined and uncorrected P values presented, increasing the possibility of type I errors. Replication in an independent dataset is required.

Moving towards meaningful measurement: Rasch analysis of the North Star Ambulatory Assessment in Duchenne muscular dystrophy

Reliable measurement of disease progression and the effect of therapeutic interventions in Duchenne muscular dystrophy (DMD) require clinically meaningful and scientifically sound rating scales. Therefore, we need robust evidence to support such tools. The North Star Ambulatory Assessment (NSAA) is a promising, clinician-rated scale with potential uses spanning clinical practice and clinical trials. In this study, we used Rasch analysis to test its suitability in these roles as a measurement instrument. NSAA data from 191 ambulant boys (mean age at assessment 7 y 8 mo, SD 2 y 4 mo; range 3 y 6 mo-15 y 5 mo) with a confirmed diagnosis of DMD were examined for psychometric properties including clinical meaning, targeting, response categories, model fit, reliability, dependency, stability, and raw to interval-level measurement. All analyses were performed using the Rasch Unidimensional Measurement Model. Overall, Rasch analysis supported the NSAA as being a reliable (high Person Separation Index of 0.91) and valid (good targeting, little misfit, no reversed thresholds) measure of ambulatory function in DMD. One item displayed misfit (lifts head, fit residual 6.9) and there was evidence for some local dependency (stand on right/left leg, climb and descend box step right/left leg, and hop on right/left leg, residual correlations >0.40), which we provide potential solutions for in future use of the NSAA. Importantly, our findings supported good clinical validity in that the hierarchy of items within the scale produced by the analyses was supported by clinical opinion, thus increasing the clinical interpretability of scale scores. In general, Rasch analysis supported the NSAA as a psychometrically robust scale for use in DMD clinical research and trials. This study also demonstrates how Rasch analysis is a useful instrument to detect and understand the key measurement issues of rating scales.

Paclitaxel, ifosfamide, and nedaplatin as second‐line treatment for patients with metastatic urothelial carcinoma: A phase II study of the SUOC group

There is no standard second-line chemotherapy treatment for recurrent or metastatic urothelial cancer (MUC). The purpose of this phase II study was to evaluate the efficacy and toxicity of the three-drug combination of paclitaxel, ifosfamide, and nedaplatin (TIN). Patients with MUC were eligible after treatment failure with methotrexate, vinblastine, doxorubicin, and cisplatin, or gemcitabine and cisplatin. Doses for TIN therapy were paclitaxel 175 mg/m2 on day 1, ifosfamide 1500 mg/m2 on days 1-3, and nedaplatin 70 mg/m2 on day 1, every 4 weeks. Tumor response, the primary efficacy parameter, was assessed according to unidimensional measurements (Response Evaluation Criteria in Solid Tumors criteria, version 1.0). Secondary efficacy parameters were overall survival (OS) and progression-free survival (PFS). Toxicity was assessed according to the National Cancer Institute Common Toxicity Criteria, version 3.0. A total of 45 patients (13 females and 32 males) with MUC were evaluable for response and toxicity. The overall response rate was 40.0%. Median PFS time was 4.0 months (95% confidence interval [CI], 4.6-11.6). Median OS time was 8.9 months (95% CI, 10.5-18.9). Grade 3 or 4 hematologic adverse events were neutropenia (95.6%), anemia (15.6%), and thrombocytopenia (17.8%). The most common grade 3 or 4 non-hematologic adverse events were anorexia (4.4%) and elevated aspartate transaminase/alanine transaminase (2.2%). No toxic death was observed. The main limitation of this study is that only 10 patients (22.2%) who were previously treated with gemcitabine and cisplatin were included. In conclusion, TIN as second-line treatment for MUC is an active regimen with a manageable toxicity profile.

Volumetry: an alternative to assess therapy response for malignant pleural mesothelioma?

The purpose of our study was to assess robustness of volumetric measurement of malignant pleural mesothelioma (MPM) before and after chemotherapy to modified RECIST (response evaluation criteria in solid tumours) criteria. 30 patients with digitally available chest computed tomography (CT) scans before and after three cycles of chemotherapy were included. Three readers independently assessed tumour response using two different methods: 1) the modified RECIST criteria; and 2) the tumour volumetric approach using dedicated software (Myrian; Intrasense, Paris, France). Inter-rater reliability of unidimensional and volumetric measurements was assessed using intraclass correlation. Tumour response classification for modified RECIST was compared to the volumetric approach applying unidimensional RECIST volumetric equivalent criteria. The determination of unidimensional tumour measurement (RECIST) revealed a low inter-rater reliability (0.55) and a low interobserver agreement for tumour response classification (general κ 0.33). Only 14 patients were classified equally. A high inter-rater reliability (0.99) and interobserver agreement (general κ 0.9) were found for absolute tumour volumes (volumetric measurements). 27 cases were classified equally. The number of cases classified as "stable disease" was higher for the volumetric approach using tumour-equivalent criteria compared to modified RECIST. Volumetric measurement of MPM on CT using Myrian software is a reliable, reproducible and sensitive method to measure tumour volume and, thus, therapy response after induction chemotherapy.

Rasch-built Overall Disability Scale (R-ODS) for immune-mediated peripheral neuropathies

To develop a patient-based, linearly weighted scale that captures activity and social participation limitations in patients with Guillain-Barré syndrome (GBS), chronic inflammatory demyelinating polyradiculoneuropathy (CIDP), and gammopathy-related polyneuropathy (MGUSP). A preliminary Rasch-built Overall Disability Scale (R-ODS) containing 146 activity and participation items was constructed, based on the WHO International Classification of Functioning, Disability and Health, literature search, and patient interviews. The preliminary R-ODS was assessed twice (interval: 2-4 weeks; test-retest reliability studies) in 294 patients who experienced GBS in the past (n = 174) or currently have stable CIDP (n = 80) or MGUSP (n = 40). Data were analyzed using the Rasch unidimensional measurement model (RUMM2020). The preliminary R-ODS did not meet the Rasch model expectations. Based on disordered thresholds, misfit statistics, item bias, and local dependency, items were systematically removed to improve the model fit, regularly controlling the class intervals and model statistics. Finally, we succeeded in constructing a 24-item scale that fulfilled all Rasch requirements. "Reading a newspaper/book" and "eating" were the 2 easiest items; "standing for hours" and "running" were the most difficult ones. Good validity and reliability were obtained. The R-ODS is a linearly weighted scale that specifically captures activity and social participation limitations in patients with GBS, CIDP, and MGUSP. Compared to the Overall Disability Sum Score, the R-ODS represents a wider range of item difficulties, thereby better targeting patients with different ability levels. If responsive, the R-ODS will be valuable for future clinical trials and follow-up studies in these conditions.

Computer-Aided Evaluation of Breast MRI for the Residual Tumor Extent and Response Monitoring in Breast Cancer Patients Receiving Neoadjuvant Chemotherapy

ObjectiveTo evaluate the accuracy of a computer-aided evaluation program (CAE) of breast MRI for the assessment of residual tumor extent and response monitoring in breast cancer patients receiving neoadjuvant chemotherapy.Materials and MethodsFifty-seven patients with breast cancers who underwent neoadjuvant chemotherapy before surgery and dynamic contrast enhanced MRI before and after chemotherapy were included as part of this study. For the assessment of residual tumor extent after completion of chemotherapy, the mean tumor diameters measured by radiologists and CAE were compared to those on histopathology using a paired student t-test. Moreover, the agreement between unidimensional (1D) measurement by radiologist and histopathological size or 1D measurement by CAE and histopathological size was assessed using the Bland-Altman method. For chemotherapy monitoring, we evaluated tumor response through the change in the 1D diameter by a radiologist and CAE and three-dimensional (3D) volumetric change by CAE based on Response Evaluation Criteria in Solid Tumors (RECIST). Agreement between the 1D response by the radiologist versus the 1D response by CAE as well as by the 3D response by CAE were evaluated using weighted kappa (k) statistics.ResultsFor the assessment of residual tumor extent after chemotherapy, the mean tumor diameter measured by radiologists (2.0 ± 1.7 cm) was significantly smaller than the mean histological diameter (2.6 ± 2.3 cm) (p = 0.01), whereas, no significant difference was found between the CAE measurements (mean = 2.2 ± 2.0 cm) and histological diameter (p = 0.19). The mean difference between the 1D measurement by the radiologist and histopathology was 0.6 cm (95% confidence interval: -3.0, 4.3), whereas the difference between CAE and histopathology was 0.4 cm (95% confidence interval: -3.9, 4.7). For the monitoring of response to chemotherapy, the 1D measurement by the radiologist and CAE showed a fair agreement (k = 0.358), while the 1D measurement by the radiologist and 3D measurement by CAE showed poor agreement (k = 0.106).ConclusionCAE for breast MRI is sufficiently accurate for the assessment of residual tumor extent in breast cancer patients receiving neoadjuvant chemotherapy. However, for the assessment of response to chemotherapy, the assessment by the radiologist and CAE showed a fair to poor agreement.

Measuring quality of life in psoriasis: the CALIPSO questionnaire

In dermatology, several instruments are available to measure health-related quality of life (HRQoL) of patients. They vary in content emphasis and in the extent to which their measurement properties have been investigated and reported. In this study, we summarized five dermatology- and psoriasis-specific HRQoL questionnaires into a new psoriasis-specific questionnaire, the Comprehensive Appraisal of Life Impact of PSOriasis (CALIPSO). The questionnaires were completed by 936 patients with a diagnosis of psoriasis. Rasch's unidimensional measurement model was used to verify the construct validity and the operational characteristics of each instrument separately. The items with the best performance according to the Rasch model were selected. The final 30 items constitute the CALIPSO. Responses are given on a 3-point scale: "never", "sometimes/quite frequently", "often/all the time". The properties of the CALIPSO now need to be verified through the analysis of new data from different populations of people with psoriasis. In future, clinicians may use it without having to decide amongst several competing scales, to evaluate HRQoL in patients with psoriasis.

The MCSS-26©: Revision of the Manchester Clinical Supervision Scale© Using the Rasch Measurement Model

Previously published accounts of the evaluation of the effects of clinical supervision, a structured system to support health service staff, have been mainly contained to small scale qualitative studies. Over the past decade, the 36-item Manchester Clinical Supervision Scale (MCSS) has transformed the evaluation landscape and has been used as a quantitative outcome measure in upward of 90 licensed studies in 12 countries worldwide. The factor structure has been replicated by other researchers and the psychometric properties have been found robust. However, it had not been previously tested empirically using newly available and sophisticated statistical analyses. This study tested the original factor structure and response format of the MCSS for goodness of fit to the Rasch model, using Rasch Unidimensional Measurement Model (RUMM) 2030 software, and investigated the validity of the questionnaire for both nursing and allied health (AH) staff. A series of Rasch analyses were conducted on the seven subscales of the MCSS. The default procedure for RUMM software uses the partial credit model, which allows items to have varying numbers of response categories and does not assume the distance between response thresholds is uniform. Detailed Rasch analyses indicated that the 36-item version of the MCSS could be reduced to 26 items and result in improved fit statistics for six subscales rather than seven. This study reconfirmed the established psychometric properties of the MCSS, now renamed the MCSS-26.

CT Tumor Volume Measurement in Advanced Non-small-cell Lung Cancer: Performance Characteristics of an Emerging Clinical Tool

Determine inter- and intraobserver variability of computed tomography (CT) tumor volume measurements in advanced non-small-cell lung cancer (NSCLC) patients treated in a Phase II clinical trial using chest CT. Twenty-three advanced NSCLC patients with a total of 53 measurable lung lesions enrolled in a Phase II, multicenter, open-label clinical trial of erlotinib were retrospectively studied with institutional review board approval. Two radiologists independently measured the tumor size, volume, and CT attenuation coefficient using commercially available volume analysis software. Concordance correlation coefficients (CCCs) and Bland-Altman plots were used to assess inter- and intraobserver agreement. High CCCs (0.949-0.990) were observed in all types of measurements for interobserver agreement. The 95% limits of agreements for volume, unidimensional, and bidimensional measurements were (-26.0%, 18.6%), (-23.1%, 24.4%), and (-34.0%, 48.6%), respectively. Volume measurement had slightly higher CCC and narrower 95% limits of agreement compared to uni- and bidimensional measurements. CCCs for intraobserver agreement were high (range, 0.946-0.996) with CCC for volume being slightly higher than CCCs of uni- and bidimensional measurements. The smaller the tumor volume was, the larger the interobserver difference of CT attenuation. Location, morphology, or adjacent atelectasis had no significant impact on inter- or intraobserver variability. CT tumor volume measurement in advanced NSCLC patients using clinical chest CT and commercially available software demonstrated high inter- and intraobserver agreement, indicating that the method may be used routinely in clinical practice.

Maximum Diameter Measurements of Aortic Aneurysms on Axial CT Images After Endovascular Aneurysm Repair: Sufficient for Follow-up?

To assess the accuracy of maximum diameter measurements of aortic aneurysms after endovascular aneurysm repair (EVAR) on axial computed tomographic (CT) images in comparison to maximum diameter measurements perpendicular to the intravascular centerline for follow-up by using three-dimensional (3D) volume measurements as the reference standard. Forty-nine consecutive patients (73 ± 7.5 years, range 51-88 years), who underwent EVAR of an infrarenal aortic aneurysm were retrospectively included. Two blinded readers twice independently measured the maximum aneurysm diameter on axial CT images performed at discharge, and at 1 and 2 years after intervention. The maximum diameter perpendicular to the centerline was automatically measured. Volumes of the aortic aneurysms were calculated by dedicated semiautomated 3D segmentation software (3surgery, 3mensio, the Netherlands). Changes in diameter of 0.5 cm and in volume of 10% were considered clinically significant. Intra- and interobserver agreements were calculated by intraclass correlations (ICC) in a random effects analysis of variance. The two unidimensional measurement methods were correlated to the reference standard. Intra- and interobserver agreements for maximum aneurysm diameter measurements were excellent (ICC = 0.98 and ICC = 0.96, respectively). There was an excellent correlation between maximum aneurysm diameters measured on axial CT images and 3D volume measurements (r = 0.93, P < 0.001) as well as between maximum diameter measurements perpendicular to the centerline and 3D volume measurements (r = 0.93, P < 0.001). Measurements of maximum aneurysm diameters on axial CT images are an accurate, reliable, and robust method for follow-up after EVAR and can be used in daily routine.

Distinguishing between Consistency and Error in Reliability Coefficients: Improving the Estimation and Interpretation of Information on Measurement Precision

Reliability coefficients indicate the proportion of total variance due to differences not attributable to error among measures separated along a quantitative continuum by a testing, survey, or assessment instrument. Person score or measure reliability is influenced by both the internal consistency of a data set and the number of items, though textbooks and research papers rarely evaluate the extent to which these factors independently affect the data in question. Probabilistic formulations of the requirements for unidimensional measurement separate consistency from error by modeling individual response processes instead of group-level variation. The utility of this separation is illustrated via analyses of small sets of simulated data, and of subsets of data from a 78-item survey of over 2,500 parents of children with disabilities. Measurement reliability ultimately concerns the structural invariance specified in models requiring sufficient statistics, parameter separation, unidimensionality, and other qualities that historically have made quantification simple, practical, and convenient for end users. The paper concludes with suggestions for a research program aimed at focusing measurement research more on the calibration and wide dissemination of tools applicable to individuals, and less on the statistical study of inter-variable relations in large data sets.

A Pilot Study of Volume Measurement as a Method of Tumor Response Evaluation to Aid Biomarker Development

Tissue biomarker discovery is potentially limited by conventional tumor measurement techniques, which have an uncertain ability to accurately distinguish sensitive and resistant tumors. Semiautomated volumetric measurement of computed tomography imaging has the potential to more accurately capture tumor growth dynamics, allowing for more exact separation of sensitive and resistant tumors and a more accurate comparison of tissue characteristics. Forty-eight patients with early stage non-small cell lung cancer and clinical characteristics of sensitivity to gefitinib were studied. High-resolution computed tomography was done at baseline and after 3 weeks of gefitinib. Tumors were then resected and molecularly profiled. Unidimensional and volumetric measurements were done using a semiautomated algorithm. Measurement changes were evaluated for their ability to differentiate tumors with and without sensitizing mutations. Forty-four percent of tumors had epidermal growth factor receptor-sensitizing mutations. Receiver operating characteristic curve analysis showed that volumetric measurement had a higher area under the curve than unidimensional measurement for identifying tumors harboring sensitizing mutations (P = 0.009). Tumor volume decrease of >24.9% was the imaging criteria best able to classify tumors with and without sensitizing mutations (sensitivity, 90%; specificity, 89%). Volumetric tumor measurement was better than unidimensional tumor measurement at distinguishing tumors based on presence or absence of a sensitizing mutation. Use of volume-based response assessment for the development of tissue biomarkers could reduce contamination between sensitive and resistant tumor populations, improving our ability to identify meaningful predictors of sensitivity.