- Email: [email protected]
Are adults with developmental disabilities more likely to visit EDs?
Correspondence
Reference [1] Cheng HH, Yen PC. Killip classification and glucose level in patients with acute myocardial infarction. Am J Emerg Med 2010;28:853-6.
Are adults with developmental disabilities more likely to visit EDs?☆ To the Editor, This brief report presents the first North American population data on rates of emergency department (ED) use among adults with intellectual and developmental disabilities (IDD) relative to the general population. Individuals with IDD are at greater risk for health problems compared with the general population [1,2] and have higher morbidity and earlier mortality as a result. Illness in individuals with IDD can be further exacerbated in a health care system that is fragmented and inadequately resourced, particularly when medical professionals do not have the necessary expertise [3]. These issues are magnified when, in addition to chronic and acute medical concerns, a comorbid mental health issue emerges. Consequently, individuals with coexisting IDD and psychiatric disorder have been identified as a group with particularly complex service needs [4]. Perhaps one of the most problematic settings for individuals with IDD and psychiatric disorder is the hospital ED. In addition to their difficulties communicating their health issues and navigating the health system [1], they may need to contend with emergency staff who feel untrained and uncomfortable working with them [5,6]. Only 2 studies have examined patterns of ED use in samples of adults with IDD relative to the general population [7,8], but both have focused on medical (nonpsychiatric) emergencies. A third study on high ED users identified IDD as a predictor of frequent psychiatric visits [9]. Together, these studies demonstrate that, with or without psychiatric issues, individuals with IDD use the ED differently than those without IDD. The current study examines ED services used by the population of adult Ontarians with IDD, with and without psychiatric disorder, relative to Ontarians without IDD. Until now, this issue could not be studied because the presence of IDD was not systematically recorded in existing ED data.
☆ This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC), as well as by annual funding from MOHLTC to the Centre for Addiction and Mental Health. The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred.
463 However, through linkage of several health data holdings (eg, physician billing data, hospitalization data) at the Institute for Clinical Evaluative Sciences in Ontario, Canada, we were able to create a service-based IDD cohort with and without psychiatric disorder, using a modified version of a previously described algorithm [10] to evaluate their ED patterns. In brief, persons were included in the IDD cohort if they had 1 of the following diagnoses: pervasive developmental disorders, mental retardation, fetal alcohol syndrome, and chromosomal and congenital anomalies for which an intellectual or developmental disability is typically present (eg, Down syndrome). Emergency department visits for 2 years (fiscal years 2007/2008 and 2008/2009) were identified using the National Ambulatory Care Reporting System. Visit rates were compared for 2 IDD groups (those with and without a psychiatric disorder) as well as for 2 groups with no IDD (those with psychiatric disorder only and a random sample of the general population). We also examined ED triage level (ie, semiurgent and nonurgent vs emergent and urgent visits) [11] and after-hour vs regular-hour visits as indirect measures of the accessibility or adequacy of non-ED types of care. In addition, the percentage of “high users,” people with 5 or more visits in 2 years, was compared across the 4 groups. Both IDD groups had consistently higher rates of ED visits when compared with the non-IDD groups (Table 1). The ratio of semiurgent and nonurgent to emergent and urgent triage visits was similar across groups. After-hour visits were common for all groups, but the ratio of after-hour to regular-hour visits was largest for those with IDD and psychiatric disorder. The proportion of high ED users was lowest in the general population sample. Among the remaining 3 groups, the highest proportion (15.6 for individuals with both IDD and a psychiatric disorder) was 3 times higher than the lowest (5.2 for those with a psychiatric disorder only). One limitation of this study is that it is based on data collected for administrative rather than research purposes. Although the diagnostic codes used to generate the 2 IDD samples have been used in previous research [10], they have not been validated against a reference standard. This study underscores the importance of gaining a detailed understanding of ED use by people with IDD, particularly those with comorbid psychiatric issues. The comparatively frequent use by individuals with IDD of a health care resource that is both expensive and difficult for them to negotiate suggests that non–emergency health care alternatives are insufficient. Future research should examine the quality of care they receive in EDs as well as their broader health care use patterns to ascertain adequacy of health care use outside EDs. Appropriate health care before, and in response to, ED visits is essential to prevent further ED visits [12]. In addition, there is a need to examine the intersection between health and social services provided to this population. Gaps in basic social services may be
464
Correspondence
Table 1 Patient demographics as of April 1, 2005, and ED visits among Ontarians during April 1, 2007, to March 31, 2009, by IDD and psychiatric disorder status IDD
Patients, n Demographics Age (y), mean (SD) Age (y), n (%) 18-24 25-34 35-44 45-54 55-64 Male (n, %) ED use Planned vs unplanned Unplanned visits, mean (SD) Planned visits, mean (SD) Ratio of unplanned to planned visits Triage level Visits for semiurgent and nonurgent reasons, mean (SD) Visits for emergent and urgent reasons, mean (SD) Ratio, (semiurgent and nonurgent)/(emergent and urgent) Time of visit After-hour ED visits, mean (SD) In-hour ED visits, mean (SD) Ratio of after-hour to in-hour (Mon-Fri, 9AM to 5PM) ED visits Frequency of visits Proportion of study patients by no. of ED visits made 0 1 to 3 4 to 6 7 to 9 10+ Percentage who are high-users (5+ visits)
No IDD
With psychiatric disorder
Without psychiatric disorder
Psychiatric disorder only
General population sample
19 759
23 790
2 042 805
1 976 469
38.51 (12.77)
37.21 (12.93)
41.41 (12.37)
39.97 (12.70)
3727 (18.9) 4206 (21.3) 4990 (25.3) 4290 (21.7) 2546 (12.9) 10 545 (53.4)
5393 (22.7) 5309 (22.3) 5526 (23.2) 4853 (20.4) 2709 (11.4) 13 683 (57.5)
237 496 397 170 535 149 514 307 358 683 808 356
293 442 (14.9) 425 153 (21.5) 499 675 (25.3) 441 942 (22.4) 316 257 (16.0) 1 043 538 (52.8)
2.87 (8.4) 0.01 (0.3) 287
1.42 (3.3) 0.01 (0.3) 142
1.08 (2.8) 0.01 (0.4) 108
0.55 (1.4) 0.01 (0.2) 55
1.56 (4.9) 1.31 (4.9) 1.19
0.75 (2.1) 0.67 (1.8) 1.12
0.57 (1.7) 0.50 (1.7) 1.14
0.26 (0.8) 0.29 (0.9) 0.90
1.89 (6.6) 0.99 (2.5) 1.91
0.86 (2.2) 0.55 (1.4) 1.56
0.68 (2.0) 0.40 (1.1) 1.7
0.35 (1.0) 0.20 (0.6) 1.75
44.6 35.8 9.3 4.1 6.4
55.9 32.3 7.1 2.5 2.2
58.9 33.3 5.1 1.4 1.3
71.8 25 2.4 0.5 0.3
15.6
8.5
5.2
1.9
another precipitant of crisis leading to emergency visits to hospital. Yona Lunsky PhD Elizabeth Lin PhD Rob Balogh PhD Centre for Addiction and Mental Health Toronto, Ontario, Canada M5V 2B4 Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5 E-mail address: [email protected] Julie Klein-Geltink MHSc Jennifer Bennie MEd Andrew S. Wilton MSc Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5
(11.6) (19.4) (26.2) (25.2) (17.6) (39.6)
Paul Kurdyak MD, PhD Centre for Addiction and Mental Health Toronto, Ontario, Canada M5V 2B4 Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5 doi:10.1016/j.ajem.2010.12.028
References [1] US Public Health Services. Closing the gap: a national blueprint to improve the health of persons with mental retardation. Report of the Surgeon General's Conference on Health Disparities and Mental Retardation. February 2001. Washington DC; 2002. [2] Van Schrojenstein Lantman-deValk HMJ, Metsemakers JFM, Haveman MJ, et al. Health problems in people with intellectual disability in general practice: a comparative study. Fam Pract 2000;17:405-7.
Correspondence [3] Ouellette-Kuntz H. Understanding health disparities and inequities faced by individuals with intellectual disabilities. J Appl Res Intellect 2005;18:113-21. [4] Lunsky Y, Garcin N, Morin D, et al. Mental health services for individuals with intellectual disabilities in Canada: findings from a national survey. J Appl Res Intellect 2007;20:439-47. [5] Mitchell R, Pearce M, Feng C, et al. ED's unequipped for all psychiatric patients: survey. Aujourd’hui 2005:7-8. [6] Lunsky Y, Gracey C, Gelfand S, et al. Emergency psychiatric services for individuals with intellectual disabilities: perspectives of hospital staff. Intellect Dev Disabil 2008;46(6):446-55. [7] Lin JD, Yen CF, Loh CH, et al. A cross-sectional study of the characteristics and determinants of emergency care utilization among people with intellectual disabilities in Taiwan. Res Dev Disabil 2006; 27(6):657-67. [8] Vencat A, Pastin RB, Hegde GG, et al. An analysis of ED utilization by adults with intellectual disability. Am J Emerg Med In press. [9] Pasic J, Russo J, Roy-Byrne P. High utilizers of emergency services. Psychiatr Serv 2005;56:678-84. [10] Balogh RS, Hunter D, Ouellette-Kuntz H. Hospital utilization among persons with an intellectual disability, Ontario, Canada, 1995-2001. J Appl Res Intellect 2005;18:181-90. [11] Physicians CAoE. Canadian Triage and Acuity Scale; 2010; http://www. caep.ca/template.asp?id=B795164082374289BBD9C1C2BF4B8D32. [12] Wood D, Hall A, Hou T, et al. Continuity of care to prevent emergency room use among persons with intellectual and developmental disabilities. J Policy Pract Intellect Disabil 2007;4(4):219-28.
Interobserver agreement in the diagnosis of acute pulmonary embolism from computed tomography pulmonary angiography and on the effectiveness of computer-aided diagnosis To the Editor, We read with great interest the article by Costantino et al [1] and their response [2] to the correspondence by ChartrandLefebvre [3] regarding interobserver agreement in the interpretation of computed tomography (CT) pulmonary angiography (CTPA) for the diagnosis of acute pulmonary embolism (PE). The interobserver agreement among radiol-
465 ogists is high for the diagnosis of massive (ie, large central) PE but is diminished for the diagnosis of segmental and subsegmental PEs. A similar hierarchy of agreement has been demonstrated for conventional pulmonary angiography [4,5]. The observations of Costantino et al are convincing and reflect the reality of interpreting CTPA examinations. CTPA is an outstanding diagnostic tool for patients suspected of PE, but the accuracy of the interpretation of CTPA depends on the radiologist's training and expertise; it has been shown that a lack of dedicated experience in CTPA interpretation results in poorer interpretative performance relative to expert CT interpreters [6,7]. Outside academic practice, general radiologists are frequently called upon to interpret CTPA studies on an emergency basis. It is therefore likely that false-positive and false-negative CTPA interpretations are not infrequent in clinical practice, as illustrated in Fig. 1. The potential for overlooked patients with PE or unnecessary anticoagulation in patients caused by falsepositive interpretation of CTPA is increased with the rapid proliferation of this technology [8,9]. These data are difficult to quantify because CTPA has been accepted as the gold standard for the diagnosis of PE, and catheter pulmonary angiograms are rarely performed for the diagnostic confirmation or exclusion of PE. Accordingly, incorrectly interpreted CTPA examinations are discovered only when CTPA studies are presented to an expert thoracic radiologist for review. As noted by Costantino et al [1,2] and ChartrandLefebvre [3], CTPA interpretation by a “second expert” would be useful in the case of nonmassive PE. However, the number of patients undergoing CTPA examinations has increased by an order of magnitude over the past decade, and the rate of positive CTPA studies is only 5% to 10% [10,11] in clinical practice; therefore, it may not be feasible to have expert CT radiologists review all negative or nonmassive PE CTPA diagnoses in clinical practice. Computer-aided diagnosis (CAD) is a nascent field [12,13], but it promises to improve the efficiency and accuracy of the
Fig. 1 Examples of false-positive (A-C) and false-negative (D) interpretations for PE on CTPA by general radiologists. The examination was ordered from the emergency department (patient A) and the inpatient units (patients B, C, and D). Examinations A, B, and C were initially interpreted as positive (arrows), and patient D was interpreted as negative for PE by general radiologists. Anticoagulation was begun for patients A, B, and C, and withheld for patient D. An experienced pulmonologist was consulted who then found patients A, B, and C were at low risk but patient D was at high risk for PE. Review of these examinations by an experienced thoracic radiologist found no evidence of PE for patients A, B, and C but confirmed PE for patient D in the left pulmonary artery (arrows). The finding mistaken for PE by general radiologists was a lymph node (patient A) and motion artifact for patients B and C.
Reference [1] Cheng HH, Yen PC. Killip classification and glucose level in patients with acute myocardial infarction. Am J Emerg Med 2010;28:853-6.
Are adults with developmental disabilities more likely to visit EDs?☆ To the Editor, This brief report presents the first North American population data on rates of emergency department (ED) use among adults with intellectual and developmental disabilities (IDD) relative to the general population. Individuals with IDD are at greater risk for health problems compared with the general population [1,2] and have higher morbidity and earlier mortality as a result. Illness in individuals with IDD can be further exacerbated in a health care system that is fragmented and inadequately resourced, particularly when medical professionals do not have the necessary expertise [3]. These issues are magnified when, in addition to chronic and acute medical concerns, a comorbid mental health issue emerges. Consequently, individuals with coexisting IDD and psychiatric disorder have been identified as a group with particularly complex service needs [4]. Perhaps one of the most problematic settings for individuals with IDD and psychiatric disorder is the hospital ED. In addition to their difficulties communicating their health issues and navigating the health system [1], they may need to contend with emergency staff who feel untrained and uncomfortable working with them [5,6]. Only 2 studies have examined patterns of ED use in samples of adults with IDD relative to the general population [7,8], but both have focused on medical (nonpsychiatric) emergencies. A third study on high ED users identified IDD as a predictor of frequent psychiatric visits [9]. Together, these studies demonstrate that, with or without psychiatric issues, individuals with IDD use the ED differently than those without IDD. The current study examines ED services used by the population of adult Ontarians with IDD, with and without psychiatric disorder, relative to Ontarians without IDD. Until now, this issue could not be studied because the presence of IDD was not systematically recorded in existing ED data.
☆ This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC), as well as by annual funding from MOHLTC to the Centre for Addiction and Mental Health. The opinions, results and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred.
463 However, through linkage of several health data holdings (eg, physician billing data, hospitalization data) at the Institute for Clinical Evaluative Sciences in Ontario, Canada, we were able to create a service-based IDD cohort with and without psychiatric disorder, using a modified version of a previously described algorithm [10] to evaluate their ED patterns. In brief, persons were included in the IDD cohort if they had 1 of the following diagnoses: pervasive developmental disorders, mental retardation, fetal alcohol syndrome, and chromosomal and congenital anomalies for which an intellectual or developmental disability is typically present (eg, Down syndrome). Emergency department visits for 2 years (fiscal years 2007/2008 and 2008/2009) were identified using the National Ambulatory Care Reporting System. Visit rates were compared for 2 IDD groups (those with and without a psychiatric disorder) as well as for 2 groups with no IDD (those with psychiatric disorder only and a random sample of the general population). We also examined ED triage level (ie, semiurgent and nonurgent vs emergent and urgent visits) [11] and after-hour vs regular-hour visits as indirect measures of the accessibility or adequacy of non-ED types of care. In addition, the percentage of “high users,” people with 5 or more visits in 2 years, was compared across the 4 groups. Both IDD groups had consistently higher rates of ED visits when compared with the non-IDD groups (Table 1). The ratio of semiurgent and nonurgent to emergent and urgent triage visits was similar across groups. After-hour visits were common for all groups, but the ratio of after-hour to regular-hour visits was largest for those with IDD and psychiatric disorder. The proportion of high ED users was lowest in the general population sample. Among the remaining 3 groups, the highest proportion (15.6 for individuals with both IDD and a psychiatric disorder) was 3 times higher than the lowest (5.2 for those with a psychiatric disorder only). One limitation of this study is that it is based on data collected for administrative rather than research purposes. Although the diagnostic codes used to generate the 2 IDD samples have been used in previous research [10], they have not been validated against a reference standard. This study underscores the importance of gaining a detailed understanding of ED use by people with IDD, particularly those with comorbid psychiatric issues. The comparatively frequent use by individuals with IDD of a health care resource that is both expensive and difficult for them to negotiate suggests that non–emergency health care alternatives are insufficient. Future research should examine the quality of care they receive in EDs as well as their broader health care use patterns to ascertain adequacy of health care use outside EDs. Appropriate health care before, and in response to, ED visits is essential to prevent further ED visits [12]. In addition, there is a need to examine the intersection between health and social services provided to this population. Gaps in basic social services may be
464
Correspondence
Table 1 Patient demographics as of April 1, 2005, and ED visits among Ontarians during April 1, 2007, to March 31, 2009, by IDD and psychiatric disorder status IDD
Patients, n Demographics Age (y), mean (SD) Age (y), n (%) 18-24 25-34 35-44 45-54 55-64 Male (n, %) ED use Planned vs unplanned Unplanned visits, mean (SD) Planned visits, mean (SD) Ratio of unplanned to planned visits Triage level Visits for semiurgent and nonurgent reasons, mean (SD) Visits for emergent and urgent reasons, mean (SD) Ratio, (semiurgent and nonurgent)/(emergent and urgent) Time of visit After-hour ED visits, mean (SD) In-hour ED visits, mean (SD) Ratio of after-hour to in-hour (Mon-Fri, 9AM to 5PM) ED visits Frequency of visits Proportion of study patients by no. of ED visits made 0 1 to 3 4 to 6 7 to 9 10+ Percentage who are high-users (5+ visits)
No IDD
With psychiatric disorder
Without psychiatric disorder
Psychiatric disorder only
General population sample
19 759
23 790
2 042 805
1 976 469
38.51 (12.77)
37.21 (12.93)
41.41 (12.37)
39.97 (12.70)
3727 (18.9) 4206 (21.3) 4990 (25.3) 4290 (21.7) 2546 (12.9) 10 545 (53.4)
5393 (22.7) 5309 (22.3) 5526 (23.2) 4853 (20.4) 2709 (11.4) 13 683 (57.5)
237 496 397 170 535 149 514 307 358 683 808 356
293 442 (14.9) 425 153 (21.5) 499 675 (25.3) 441 942 (22.4) 316 257 (16.0) 1 043 538 (52.8)
2.87 (8.4) 0.01 (0.3) 287
1.42 (3.3) 0.01 (0.3) 142
1.08 (2.8) 0.01 (0.4) 108
0.55 (1.4) 0.01 (0.2) 55
1.56 (4.9) 1.31 (4.9) 1.19
0.75 (2.1) 0.67 (1.8) 1.12
0.57 (1.7) 0.50 (1.7) 1.14
0.26 (0.8) 0.29 (0.9) 0.90
1.89 (6.6) 0.99 (2.5) 1.91
0.86 (2.2) 0.55 (1.4) 1.56
0.68 (2.0) 0.40 (1.1) 1.7
0.35 (1.0) 0.20 (0.6) 1.75
44.6 35.8 9.3 4.1 6.4
55.9 32.3 7.1 2.5 2.2
58.9 33.3 5.1 1.4 1.3
71.8 25 2.4 0.5 0.3
15.6
8.5
5.2
1.9
another precipitant of crisis leading to emergency visits to hospital. Yona Lunsky PhD Elizabeth Lin PhD Rob Balogh PhD Centre for Addiction and Mental Health Toronto, Ontario, Canada M5V 2B4 Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5 E-mail address: [email protected] Julie Klein-Geltink MHSc Jennifer Bennie MEd Andrew S. Wilton MSc Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5
(11.6) (19.4) (26.2) (25.2) (17.6) (39.6)
Paul Kurdyak MD, PhD Centre for Addiction and Mental Health Toronto, Ontario, Canada M5V 2B4 Institute for Clinical Evaluative Sciences Toronto, ON, Canada M4N3M5 doi:10.1016/j.ajem.2010.12.028
References [1] US Public Health Services. Closing the gap: a national blueprint to improve the health of persons with mental retardation. Report of the Surgeon General's Conference on Health Disparities and Mental Retardation. February 2001. Washington DC; 2002. [2] Van Schrojenstein Lantman-deValk HMJ, Metsemakers JFM, Haveman MJ, et al. Health problems in people with intellectual disability in general practice: a comparative study. Fam Pract 2000;17:405-7.
Correspondence [3] Ouellette-Kuntz H. Understanding health disparities and inequities faced by individuals with intellectual disabilities. J Appl Res Intellect 2005;18:113-21. [4] Lunsky Y, Garcin N, Morin D, et al. Mental health services for individuals with intellectual disabilities in Canada: findings from a national survey. J Appl Res Intellect 2007;20:439-47. [5] Mitchell R, Pearce M, Feng C, et al. ED's unequipped for all psychiatric patients: survey. Aujourd’hui 2005:7-8. [6] Lunsky Y, Gracey C, Gelfand S, et al. Emergency psychiatric services for individuals with intellectual disabilities: perspectives of hospital staff. Intellect Dev Disabil 2008;46(6):446-55. [7] Lin JD, Yen CF, Loh CH, et al. A cross-sectional study of the characteristics and determinants of emergency care utilization among people with intellectual disabilities in Taiwan. Res Dev Disabil 2006; 27(6):657-67. [8] Vencat A, Pastin RB, Hegde GG, et al. An analysis of ED utilization by adults with intellectual disability. Am J Emerg Med In press. [9] Pasic J, Russo J, Roy-Byrne P. High utilizers of emergency services. Psychiatr Serv 2005;56:678-84. [10] Balogh RS, Hunter D, Ouellette-Kuntz H. Hospital utilization among persons with an intellectual disability, Ontario, Canada, 1995-2001. J Appl Res Intellect 2005;18:181-90. [11] Physicians CAoE. Canadian Triage and Acuity Scale; 2010; http://www. caep.ca/template.asp?id=B795164082374289BBD9C1C2BF4B8D32. [12] Wood D, Hall A, Hou T, et al. Continuity of care to prevent emergency room use among persons with intellectual and developmental disabilities. J Policy Pract Intellect Disabil 2007;4(4):219-28.
Interobserver agreement in the diagnosis of acute pulmonary embolism from computed tomography pulmonary angiography and on the effectiveness of computer-aided diagnosis To the Editor, We read with great interest the article by Costantino et al [1] and their response [2] to the correspondence by ChartrandLefebvre [3] regarding interobserver agreement in the interpretation of computed tomography (CT) pulmonary angiography (CTPA) for the diagnosis of acute pulmonary embolism (PE). The interobserver agreement among radiol-
465 ogists is high for the diagnosis of massive (ie, large central) PE but is diminished for the diagnosis of segmental and subsegmental PEs. A similar hierarchy of agreement has been demonstrated for conventional pulmonary angiography [4,5]. The observations of Costantino et al are convincing and reflect the reality of interpreting CTPA examinations. CTPA is an outstanding diagnostic tool for patients suspected of PE, but the accuracy of the interpretation of CTPA depends on the radiologist's training and expertise; it has been shown that a lack of dedicated experience in CTPA interpretation results in poorer interpretative performance relative to expert CT interpreters [6,7]. Outside academic practice, general radiologists are frequently called upon to interpret CTPA studies on an emergency basis. It is therefore likely that false-positive and false-negative CTPA interpretations are not infrequent in clinical practice, as illustrated in Fig. 1. The potential for overlooked patients with PE or unnecessary anticoagulation in patients caused by falsepositive interpretation of CTPA is increased with the rapid proliferation of this technology [8,9]. These data are difficult to quantify because CTPA has been accepted as the gold standard for the diagnosis of PE, and catheter pulmonary angiograms are rarely performed for the diagnostic confirmation or exclusion of PE. Accordingly, incorrectly interpreted CTPA examinations are discovered only when CTPA studies are presented to an expert thoracic radiologist for review. As noted by Costantino et al [1,2] and ChartrandLefebvre [3], CTPA interpretation by a “second expert” would be useful in the case of nonmassive PE. However, the number of patients undergoing CTPA examinations has increased by an order of magnitude over the past decade, and the rate of positive CTPA studies is only 5% to 10% [10,11] in clinical practice; therefore, it may not be feasible to have expert CT radiologists review all negative or nonmassive PE CTPA diagnoses in clinical practice. Computer-aided diagnosis (CAD) is a nascent field [12,13], but it promises to improve the efficiency and accuracy of the
Fig. 1 Examples of false-positive (A-C) and false-negative (D) interpretations for PE on CTPA by general radiologists. The examination was ordered from the emergency department (patient A) and the inpatient units (patients B, C, and D). Examinations A, B, and C were initially interpreted as positive (arrows), and patient D was interpreted as negative for PE by general radiologists. Anticoagulation was begun for patients A, B, and C, and withheld for patient D. An experienced pulmonologist was consulted who then found patients A, B, and C were at low risk but patient D was at high risk for PE. Review of these examinations by an experienced thoracic radiologist found no evidence of PE for patients A, B, and C but confirmed PE for patient D in the left pulmonary artery (arrows). The finding mistaken for PE by general radiologists was a lymph node (patient A) and motion artifact for patients B and C.