Outcome of Recommendations for Radiographic Follow-Up of Pneumonia on Outpatient Chest Radiography

Transcription

1 Cardiopulmonary Imaging Original Research Little et al. Radiographic Follow-Up of Pneumonia Cardiopulmonary Imaging Original Research Brent P. Little 1 Matthew D. Gilman 2 Kathryn L. Humphrey 2 Tarik K. Alkasab 2 Fiona K. Gibbons 3 Jo-Anne O. Shepard 2 Carol C. Wu 2 Little BP, Gilman MD, Humphrey KL, et al. Keywords: chest radiography, community-acquired pneumonia, lung cancer DOI: /AJR Received March 8, 2013; accepted after revision June 16, Department of Radiology and Imaging Sciences, Emory University, Atlanta, GA. 2 Department of Radiology, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St., Boston, MA Address correspondence to C. C. Wu (carolcwu@gmail.com). 3 Department of Pulmonary and Critical Care Unit, Massachusetts General Hospital, Harvard Medical School. Boston, MA. This article is available for credit. AJR 2014; 202: X/14/ American Roentgen Ray Society Outcome of Recommendations for Radiographic Follow-Up of Pneumonia on Outpatient Chest Radiography OBJECTIVE. Follow-up chest radiographs are frequently recommended by radiologists to document the clearing of radiographically suspected pneumonia. However, the clinical utility of follow-up radiography is not well understood. The purpose of this study was to examine the incidence of important pulmonary pathology revealed during follow-up imaging of suspected pneumonia on outpatient chest radiography. MATERIALS AND METHODS. Reports of 29,138 outpatient chest radiography examinations performed at an academic medical center in 2008 were searched to identify cases in which the radiologist recommended follow-up chest radiography for presumed community-acquired pneumonia (n = 618). Descriptions of index radiographic abnormalities were recorded. Reports of follow-up imaging (radiography and CT) performed during the period from January 2008 to January 2010 were reviewed to assess the outcome of the index. Clinical history, demographics, microbiology, and pathology reports were reviewed and recorded. RESULTS. Compliance with follow-up imaging recommendations was 76.7%. In nine of 618 cases (1.5%), a newly diagnosed malignancy corresponded to the on chest radiography initially suspected to be pneumonia. In 23 of 618 cases (3.7%), an alternative nonmalignant disease corresponded with the on chest radiography initially suspected to be pneumonia. Therefore, in 32 of 618 patients (5.2%), significant new pulmonary diagnoses were established during follow-up imaging of suspected pneumonia. CONCLUSION. Follow-up imaging of radiographically suspected pneumonia leads to a small number of new diagnoses of malignancy and important nonmalignant diseases, which may alter patient management. M alignancy and other important nonmalignant pulmonary diseases may present with a radiographic appearance similar to pneumonia. For this reason, follow-up chest radiography is often recommended in patients to confirm resolution of a radiographically suspected pneumonia and to exclude an alternative disease (such as cancer), that would not clear after treatment of pneumonia. However, there is no consensus regarding the clinical utility of follow-up radiography. The 2005 guidelines of the American College of Chest Physicians recommend follow-up radiography approximately 8 weeks after diagnosis to ensure clearing [1]. The British Thoracic Society (2009) recommends follow-up chest radiography at 6 weeks for all patients who have persistent clinical symptoms or who are at increased risk for malignancy, such as smokers and patients older than 50 years [2]. However, the 2007 Infectious Disease Society of America and the American Thoracic Society consensus guidelines for managing community-acquired pneumonia in adults contain no mention of follow-up chest radiography [3, 4]. Clinical practice regarding follow-up recommendations of a new pulmonary opacity suspected to be pneumonia varies among radiologists. A survey of the members of the Society of Thoracic Radiology showed that 42% of respondents always recommend follow-up radiography for new pulmonary opacities suspected to represent communityacquired pneumonia. The majority of respondents (55%) recommend follow-up radiography sometimes on the basis of appearance of the opacity, patient age, and clinical symptoms [5]. The yield of such follow-up imaging recommendations and the rates of compliance are incompletely understood. The purpose of this study was to retrospectively evaluate the outcome of recommendations for follow-up radiography at our 54 AJR:202, January 2014

2 TABLE 1: Characteristics of Patient Populations Radiographic Follow-Up of Pneumonia Characteristic All Patients With Recommendations for Follow-Up Radiography (n = 805) institution, where the current practice is to routinely recommend follow-up chest radiography in all adults, regardless of age or smoking history, to confirm clearance of pulmonary opacities suspected to represent community-acquired pneumonia. Materials and Methods This retrospective HIPAA-compliant study was approved by the institutional review board. The requirement for informed consent was waived. Patient Selection Electronic records of 29,138 chest radiography examinations performed in the outpatient setting at an academic medical center during 1 year (2008) were queried for cases in which the interpreting radiologist recommended follow-up chest radiography to ensure resolution of radiographically suspected pneumonia. Using custom software written using the Ruby programming language ( accessed November 12, 2012), our department s radiology information system (Centricity, GE Healthcare) was searched for examinations with chest radiography examination codes in which the reports contained the text recommend but not the phrase no imaging recommendations. The subjects were outpatients referred from general medicine clinics, specialty clinics, and the emergency department associated with a tertiary care academic center. Index radiography examinations with findings that the interpreting radiologist thought were highly suspicious for cancer or recommended immediate chest CT (n = 1317) or PET/CT (n = 1) were excluded from the study. Patient Data Collection Patient results and clinical notes from January 2008 to January 2010 were reviewed in the electronic medical record (Partners EMR, Partners Patients With Follow-Up Imaging (n = 618) Healthcare Clinical Applications Suite, Clinical Systems Research and Development, Partners Healthcare System; and Partners Healthcare Longitudinal Medical Record, Clinical Informatics Research and Development, Partners Healthcare System). Data were recorded using Microsoft Office Excel The patient demographics are listed in Table 1. Results of cytology, surgical pathology, bronchoscopy washings, and microbiology were obtained and recorded from the Partners EMR. Final diagnoses were obtained from cytology, surgical pathology, bronchoscopy results, microbiology, and patient notes indicating a final clinical diagnosis. Radiographic and CT Interpretation Radiography and chest CT examinations in the study group were interpreted by 11 thoracic radiologists with 2 40 years of experience. Digital posteroanterior and lateral chest radiography were performed on computed radiography and digital radiography units. Chest CT examinations were performed on 16-MDCT and 64-MDCT scanners. All studies were reviewed on a dedicated PACS workstation (Impax, version 5.3.2, AGFA Healthcare). Patients With No Follow-Up Imaging (n = 187) p (Follow-Up vs No Follow-Up) Mean age (median) (y) 61 (62) 63 (64) 54 (55) < Sex (M:F) 381: :333 96: History of cancer 201 (25) 169 (27) 32 (17) Current smoker 125 (16) 99 (16) 26 (14) 0.73 Documented COPD 127 (16) 114 (18) 13 (7) < Diabetes 108 (13) 89 (14) 19 (10) 0.14 Undergoing chemotherapy 25 (3) 23 (3.7) 2 (1) Known HIV 14 (2) 12 (1.9) 2 (1) 0.54 Preoperative examination 36 (4) 22 (4) 14 (7) Note Except where indicated otherwise, data in parentheses are percentages. Bold text indicates significant difference. COPD = chronic obstructive pulmonary disease. Radiographic and CT Data Collection The dates of all radiography and chest CT examinations were recorded. Reports of all chest radiography and all chest CT examinations performed during the follow-up period (January 2008 January 2010) were reviewed, terminating with a study showing the resolution, pathologic diagnosis, or last available study performed on the patient within the hospital system. Resolution was defined as absence of a corresponding on follow-up radiography or CT. Cases in which no was found on CT or in which the nearly resolved with only minimal atelectasis or scarring were also included in this category. Initial follow-up radiographs were classified as showing either resolution of, improvement but incomplete resolution, no significant interval change, or worsening. Specific descriptions of the original and follow-up chest radiography abnormalities were recorded and are listed in Table 2. CT results were recorded, including the impression of the thoracic radiologist interpreting the examination (infection or aspiration, malignancy, subsegmental atelectasis or scarring, edema, and so on). Index radiography reports were compared TABLE 2: Description of Abnormality on Index Radiography Cases in Which Follow-Up Radiography or CT Was Performed (n = 618) Abnormality No. (%) Opacity not otherwise specified 338 (54.7) Pneumonia/consolidation 109 (17.6) Atelectasis vs pneumonia 87 (14.1) Reticular opacity/interstitial 53 (8.6) Nodular opacity or nodules 29 (4.7) Pleural effusion 11 (1.8) Rounded opacity 9 (1.5) Masslike opacity 2 (0.3) AJR:202, January

3 Little et al. with CT reports to assess anatomic correspondence between radiographic and CT abnormalities. Cases of Malignancy, Infection, and Other Nonmalignant Disease In cases in which malignancy, infection, or other nonmalignant disease was the eventual diagnosis, the original chest radiographs and chest CT images were reviewed independently by two thoracic radiologists to confirm that the malignancy, infection, or other nonmalignant disease corresponded with the originally detected radiographic (Fig. 1). One of the two thoracic radiologists did not interpret or report any of the cases originally. The other thoracic radiologist only originally reported one of the cases. These two radiologists were not blinded to the outcome. Cytology, surgical pathology, bronchoscopy, and microbiology results were recorded, and clinical notes were reviewed. Statistical Analysis Statistical analysis was performed using Prism 6 (GraphPad Software). Differences in characteristics between patient groups were assessed for statistical significance using the Fisher exact test. Calculations of significance for mean age between groups were performed using a two-sample unequal variance Student t test. A p value of less than 0.05 was considered statistically significant. Results Study Population and Description of Index Abnormalities In 2008, 29,138 outpatient chest radiography examinations were performed at our academic center. The reports of 2996 of these examinations matched the initial screen for recommendations. Our study group included 805 consecutive reports that were found to include a recommendation for radiographic follow-up to resolution of suspected pneumonia. The patients ranged in age from 20 to 98 years (424 women, 381 men; mean age, 61 years; median, 62 years); 125 (16%) of these patients were current smokers. The most common indications for the original chest radiography were cough (611, 76%), fever (314, 39%), chest pain (142, 18%), dyspnea (99, 12%), and wheezing (40, 5%). Other demographic and clinical data are summarized in Table 1. The original descriptions of index abnormalities are summarized in Table 2. Six hundred eighteen of 805 (77%) patients for whom follow-up radiography was recommended underwent documented follow-up imaging (285 men and 333 women). One hundred eighty-eight (23%) patients underwent no follow-up chest imaging (CT or radiography) at our institution. The average age of patients without follow-up imaging (mean, 54 years; median, 55 years) was lower than that of the patients with follow-up imaging (mean, 63 years; median, 64 years p < ). Patients with documented follow-up imaging (27%) were more likely to have a history of known malignancy than those without followup imaging (17% p = 0.005) and were more likely to have documented chronic obstructive pulmonary disease (COPD) (18%) than those without follow-up (7% p < ). Results of Follow-Up Imaging The results of follow-up imaging are listed in Tables 3 and 4 and are summarized in the flowchart in Figure 2. Malignancy Discovered During Follow-Up In total, 15 cases of previously undiagnosed cancer or previously unknown recurrent malignancy were found in the study group during follow-up imaging. Nine of these malignancies corresponded with the on the index radiograph. Of these nine patients, eight had a new diagnosis or a new recurrence of non small cell lung cancer. One patient had large B cell lymphoma. On chest radiography, only one of nine cases of malignancy was described as a nodule. The remaining eight cases were described as an opacity or consolidation. On chest CT for these nine patients, seven had nodules, two had masses not causing lobar or segmental atelectasis, two had consolidation, and one had a mass with associated right middle and lower lobe atelectasis. The mean age of these nine patients was 68.4 years (median, 73 years; range, years). Five of the nine had a history of malignancy at the time of the index radiography; three had known COPD, and two were current smokers. The presenting symptoms of these nine patients were cough (n = 6), chest pain (n = 2), wheezing (n = 2), dyspnea (n = 2), fatigue (n = 2), fever (n = 1), night sweats (n = 1), and weight loss (n = 1). In the six cases of malignancy not correlating with the location on the index chest radiographic, only one case was an extrathoracic malignancy (pancreatic neuroendocrine tumor without lung malignancy). The remaining cases of malignancy were thoracic metastasis (lung, colon, chondrosarcoma, head and neck tumor), thoracic involvement by chronic lymphocytic leukemia, and lymphangitic carcinomatosis (breast). Other Important Diagnoses In addition to newly diagnosed malignancy, 23 patients had important nonmalignant diagnoses made on follow-up imaging. Tuberculosis or atypical mycobacterial infection (n = 6), fungal infection (n = 5), organizing or eosinophilic pneumonia (n = 5), rounded atelectasis (n = 3), alveolar hemorrhage (n = 1), lung abscess (n = 1), Pneumocystis jiroveci (n = 1), and septic emboli (n = 1) were among the diagnoses made that corresponded with the original chest radiographic. The mean age of these 23 patients was 56 years (median, 58 years; range, years). Time to Follow-Up Imaging The mean time to first follow-up imaging (radiography or CT) was 80 days (median, 39 days; range, days). The average time between index radiography and CT, if performed, was 78 days (median, 41 days; range, days). In eight of nine patients with newly detected malignancy corresponding with the radiographic, the mean TABLE 3: Modality of Follow-Up and Time to Follow-Up (n = 805) Characteristic Value Any follow-up imaging 618 (77) No follow-up imaging 187 (23) Follow-up radiography only 432 (53.6) Follow-up radiography and CT 141 (17.5) Follow-up CT only 45 (5.6) Patients with > 1 follow-up radiography study 185 (23) Mean time to any first follow-up study 80 d (median, 39 d; range, d) Mean time to first follow-up radiography 78 d (median, 41 d; range, d) Mean time to first follow-up CT 78 d (median, 41 d; range, d) Patients with first follow-up within 90 d 480/618 (60) Note Except where indicated otherwise, data in parentheses are percentages. 56 AJR:202, January 2014

4 Radiographic Follow-Up of Pneumonia A C B D Fig year-old woman with 3 4 day history of productive cough, fever, and pleuritic chest pain. A and B, Initial posteroanterior (A) and lateral (B) chest radiographs were interpreted as lingular pneumonia (arrows) and small left pleural effusion. Follow-up chest radiography was recommended to ensure clearance. C and D, Follow-up posteroanterior (C) and lateral (D) chest radiographs obtained 7 weeks later to ensure clearing of pneumonia (with improved clinical symptoms) show incomplete clearing of lingular opacity (arrows) and enlargement of left hilum. Chest CT was recommended. E and F, Axial contrast-enhanced CT images of chest obtained 2 days later in lung (E) and mediastinal (F) windows show atelectasis in lingula (arrow, E) and central non small cell lung cancer obstructing lingular bronchi (arrow, F). E F AJR:202, January

5 Little et al. TABLE 4: Important Diagnoses Other Than Pneumonia Corresponding With Radiographic Abnormality Diagnosis No of Patients Malignancy 9 Non small cell lung cancer 8 Large B cell lymphoma 1 Important benign diagnoses 23 Tuberculosis/atypical mycobacterial infection 6 Eosinophilic or organizing pneumonia 5 Fungal infection 5 Round atelectasis 3 Abscess 1 Alveolar hemorrhage 1 Pneumocystis pneumonia 1 Septic emboli 1 Initial radiographic follow-up 573/618 (92.7%) Persistence or worsening 177/573 (30.9%) No chest CT 90/177 (50.8%) Lost to follow-up 89/90 (98.9%) Cancer not matching chest radiography 6/87 (6.9%) Patients who underwent follow-up imaging 618/805 (77%) Complete resolution 396/573 (69.1%) Cancer matching 1/13 (7.7%) Chest CT performed 87/177 (49.2%) Cancer matching chest radiography 1/90 (1.1%) Cancer matching chest radiography 7/87 (8%) Persistence or worsening 13/45 (28.9%) CT follow-up only 45/618 (7.3%) Important benign diagnosis round atelectasis 2 fungal infection 2 4/13 (30.8%) Important benign diagnosis tuberculosis/atypical mycobacteria 6 eosinophilic/organizing pneumonia 5 fungal infection 3 pneumocystitis 1 abscess 1 septic emboli 1 alveolar hemorrhage 1 round atelectasis 1 19/87 (23.8%) Fig. 2 Flowchart shows outcome of patients who underwent follow-up imaging. Resolution of or nothing corresponding with index chest radiograph 37/45 (71.1%) Other incidental 8/13 (61.5%) Other incidental 55/87 (63.2%) time to first follow-up imaging (radiography or CT) was 17.9 days (median, 16 days; range, 6 31 days). Only one patient had a significant delay, with follow-up radiography performed 186 days after the index study, CT performed 2 days subsequent to follow-up radiography, and an eventual diagnosis of moderately differentiated lung adenocarcinoma, corresponding with the original. For the 23 patients with important nonmalignant diagnoses corresponding with the radiographic, the mean time to first followup imaging (radiography or CT) was 70 days (median, 8 days; range, days). Discussion Thoracic malignancy, atypical infection, and other nonmalignant pulmonary diseases may appear similar to community-acquired pneumonia on chest radiography [6 8]. Although practice varies, many clinicians and radiologists recommend follow-up chest radiography to confirm clearing of a radiographic opacity suspected to represent communityacquired pneumonia [5]. The rationale for this practice is to ensure that the patient s symptoms and radiographic findings are in fact due to community-acquired pneumonia rather than another disease. The purpose of this study was to investigate the outcome of recommendations for follow-up chest radiography in patients with a pulmonary opacity initially thought to represent community-acquired pneumonia. Prior research investigating the radiographic follow-up of patients diagnosed with pneumonia has focused on lung cancer as the primary outcome. These studies show a wide range ( %) in the incidence of cancer in patients diagnosed with pneumonia [9 14]. The most recent study of 3398 patients in Edmonton, Alberta, Canada, found a lung cancer incidence of 2.5% within a subgroup of 1354 patients who underwent chest radiography at 90 days [13]. The highest rate of lung cancer was found in a study by Mortensen et al. [14] at 9.2%. This high incidence of lung cancer compared with other studies may be due to the characteristics of the population studied: patients of a Veterans Affairs institution, more than 65 years old, inpatients, and predominantly men. These prior studies document the epidemiologic association of pneumonia and lung cancer but do not specifically show that the chest radiograph opacity correlated directly with the lung cancer or was a direct effect of the lung cancer itself. In our study, we found nine patients (1.5%) in whom a malignancy directly corresponded 58 AJR:202, January 2014

6 with the index chest radiographic initially suspected to represent pneumonia. Our 1.5% incidence of malignancy (not initially suspected to be neoplasm on chest radiography) is similar to that found by Woodhead et al. [12] (1.7%) and Bochud et al. [10] (1.2%). Although Tang et al. [13] found a higher rate of lung malignancy (2.5%), they included all cases of lung cancer whereas our 1.5% incidence of malignancy excludes patients in whom the index radiographic did not correspond with the ultimately diagnosed cancer. If we include all cases of new malignancy regardless of whether the cancer correlated with the chest radiographic finding, our incidence of new malignancy is similar to that of Tang et al. s at 2.4%. We found that most patients diagnosed with malignancy on follow-up imaging were over the age of 50 years. Tang et al. [13] also found a higher incidence of cancer in patients over 50 years and proposed that radiographic follow-up for pneumonia should be targeted to patients older than 50 years. In our study, however, two patients (47 and 48 years old) with new diagnoses of non small cell lung cancer corresponding with the original radiographic were under the age of 50 years, accounting for 22% of the new diagnoses of malignancies corresponding with the index radiographic. In addition to lung cancer and other lung malignancies, specific alternative nonmalignant diseases, such as atypical infection, organizing pneumonia, sarcoidosis, and bronchopulmonary sequestration, may also be initially diagnosed as community-acquired pneumonia. Diagnosis of such diseases by failure to resolve on follow-up chest radiography after treatment of pneumonia can be important for appropriate patient management. To our knowledge, the diagnostic yield of follow-up chest radiography in the detection of diseases other than lung cancer has not previously been formally reported. Our study addresses this issue and investigates the diagnostic yield of follow-up chest radiography in the detection of such nonmalignant diseases. We found 23 patients (3.6%) in whom a significant nonmalignant disease (other than community-acquired pneumonia) corresponded with the index chest radiographic finding initially suspected to represent community-acquired pneumonia. When the number of alternative nonmalignant diseases and new diagnoses of malignancy are combined, follow-up chest radiography for suspected community Radiographic Follow-Up of Pneumonia acquired pneumonia allowed the detection of an alternative clinically significant disease in a total of 32 patients (5%). A limitation of our study is that we were not able to achieve 100% compliance for follow-up radiography. Our compliance for follow-up imaging was 60% at 90 days; overall, 77% of patients received follow-up imaging. This compares favorably with a recent study by Tang et al. [13], who reported 40% compliance at 90 days. In addition, we were not able to determine in all cases whether the follow-up chest radiography was performed solely to reevaluate the initial chest radiographic or if it was performed for other reasons, such as persistence of clinical symptoms. Other limitations of our study include the retrospective nature of the data gathering and a population limited to outpatients only. In our methods, we also used the radiology reports for the analysis rather than a consensus review of all original images. However, the use of radiology reports provided the opportunity to examine the outcome of real-world clinical radiology practice. Thus, our study was not designed to assess inter- or intraobserver variation, and we do not know how many resolved pneumonias actually represented a diagnostic disagreement among radiologists. Because our study was performed at one institution by dedicated thoracic radiologists, the number of patients excluded because of a suspicious radiographic finding (i.e., who were referred directly to CT) could bias the number of significant alternative diagnoses made during follow-up, potentially reducing this number. Lastly, because our study is not a randomized study comparing patients who received imaging with those who received no imaging, additional studies are needed to define the clinical benefit of routine radiography. Despite these limitations, we believe the data provide evidence that follow-up chest radiography in the setting of suspected pneumonia leads to a small number of new diagnoses of malignancy and important nonmalignant diseases for which the therapeutic approach may be altered. Our data suggest that the majority (69%) of radiographic abnormalities initially suspected to represent pneumonia would resolve on follow-up chest radiography. However, in patients with radiographically suspected pneumonia, the early detection of a small number of important alternative diagnoses may depend on the use of radiographic follow-up. References 1. Ramsdell J, Narsavage GL, Fink JB. Management of community-acquired pneumonia in the home: an American College of Chest Physicians clinical position statement. Chest 2005; 127: Lim WS, Baudouin SV, George RC, et al. BTS guidelines for the management of community acquired pneumonia in adults: update Thorax 2009; 64(suppl 3):iii1 iii55 3. Mandell LA, Wunderink RG, Anzueto A, et al. Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis 2007; 44(suppl 2):S27 S72 4. Niederman MS, Mandell LA, Anzueto A, et al.; American Thoracic Society. Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001; 163: Humphrey KL, Gilman MD, Little BP, et al. Radiographic follow-up of suspected pneumonia: survey of Society of Thoracic Radiology membership. J Thorac Imaging 2013; 28: Rome L, Murali G, Lippmann M. Nonresolving pneumonia and mimics of pneumonia. Med Clin North Am 2001; 85: Weyers CM, Leeper KV. Nonresolving pneumonia. Clin Chest Med 2005; 26: Kuru T, Lynch JP. Nonresolving or slowly resolving pneumonia. Clin Chest Med 1999; 20: Holmberg H, Kragsbjerg P. Association of pneumonia and lung cancer: the value of convalescent chest radiography and follow-up. Scand J Infect Dis 1993; 25: Bochud PY, Moser F, Erard P, et al. Communityacquired pneumonia: a prospective outpatient study. Medicine (Baltimore) 2001; 80: Marrie TJ. Pneumonia and carcinoma of the lung. J Infect 1994; 29: Woodhead MA, Macfarlane JT, McCracken JS, Rose DH, Finch RG. Prospective study of the aetiology and outcome of pneumonia in the community. Lancet 1987; 329: Tang KL, Eurich DT, Minhas-Sandhu JK, Marrie TJ, Majumdar SR. Incidence, correlates, and chest radiographic yield of new lung cancer diagnosis in 3398 patients with pneumonia. Arch Intern Med 2011; 171: Mortensen EM, Copeland LA, Pugh MJ, et al. Diagnosis of pulmonary malignancy after hospitalization for pneumonia. Am J Med 2010; 123:66 71 FOR YOUR INFORMATION This article is available for CME and Self-Assessment (SA-CME) credit that satisfies Part II requirements for maintenance of certification (MOC). To access the examination for this article, follow the prompts associated with the online version of the article. AJR:202, January