0
Original Research: Pulmonary Vascular Disease |

Correlation Between Early Direct Communication of Positive CT Pulmonary Angiography Findings and Improved Clinical OutcomesCommunication and Outcomes of Pulmonary Embolism FREE TO VIEW

Kanako K. Kumamaru, MD, PhD; Andetta R. Hunsaker, MD; Hiraku Kumamaru, MD, MPH; Elizabeth George, MBBS; Arash Bedayat, MD; Frank J. Rybicki, MD, PhD
Author and Funding Information

From the Applied Imaging Science Laboratory, Department of Radiology (Drs K. Kumamaru, Hunsaker, George, Bedayat, and Rybicki), Brigham and Women’s Hospital & Harvard Medical School; and the Department of Epidemiology (Dr H. Kumamaru), Harvard School of Public Health, Boston, MA.

Correspondence to: Frank J. Rybicki, MD, PhD, Applied Imaging Science Laboratory, Department of Radiology, Brigham and Women’s Hospital & Harvard Medical School, 75 Francis St, Boston, MA 02115; e-mail: frybicki@partners.org


Dr Bedayat is currently at the Department of Radiology, University of Massachusetts Medical School, Worcester, MA.

Part of this paper was presented at the 97th Scientific Assembly and Annual Meeting of the Radiological Society of North America, on November 27-December 2, 2011, Chicago, IL.

Funding/Support: Dr K. Kumamaru was supported by The Japan Society for the Promotion of Science, as a Postdoctoral Fellow for Research Abroad, to conduct this study.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2013;144(5):1546-1554. doi:10.1378/chest.13-0308
Text Size: A A A
Published online

Background:  Despite a general consensus that rapid communication of critical radiology findings from radiologists to referring physicians is imperative, a possible association with superior patient outcomes has not been confirmed. The objective of this study was to evaluate the correlation between early direct communication of CT image findings by radiologists to referring physicians and better clinical outcomes in patients with acute pulmonary embolism (PE).

Methods:  This was a retrospective, single-institution, cohort study that included 796 consecutive patients (February 2006 to March 2010) who had acute PE confirmed by CT pulmonary angiography (CTPA) and whose treatment had not been initiated at the time of CTPA acquisition. The time from CTPA to direct communication of the diagnosis was evaluated for its association with time from CTPA to treatment initiation and with 30-day mortality. Cox regression analysis was performed with inverse probability weighting by propensity scores calculated using 20 potential confounding factors.

Results:  In 93.4% of patients whose first treatment was anticoagulation, the referring physicians started treatment after receiving direct notification of the diagnosis from the radiologist. Late communication (> 1.5 h after CTPA; n = 291) was associated with longer time to treatment initiation (adjusted hazard ratio [HR], 0.714; 95% CI, 0.610-0.836; P < .001) and higher all-cause and PE-related 30-day mortality (HR, 1.813; 95% CI, 1.163-2.828; P = .009; and HR, 2.625; 95% CI, 1.362-5.059; P = .004, respectively).

Conclusions:  Delay (> 1.5 h of CTPA acquisition) in direct communication of acute PE diagnosis from radiologists to referring physicians was significantly correlated with a higher risk of delayed treatment initiation and death within 30 days.

Figures in this Article

Critical radiology findings “which could result in mortality or significant morbidity if appropriate diagnostic and/or therapeutic follow-up steps are not undertaken”1,2 should be rapidly communicated from radiologists to referring physicians “in a manner that reasonably ensures timely receipt of the findings.”15 However, despite general consensus of its importance and clinical benefit to patient outcomes, to our knowledge there is no published evidence that confirms the possible relationship between rapid communication and clinical outcomes.

Patients with acute pulmonary embolism (PE) are a good target population to test this general consensus because the diagnosis is usually confirmed by CT pulmonary angiography (CTPA),68 and the appropriate communication of imaging findings could, thus, be critical to patient care. As empirical anticoagulation is recommended only in specific situations,9 the treatment is initiated after confirmation of diagnosis in the majority of cases. Therefore, if communication from radiologists contributes to clinical decision-making, the time to treatment initiation would be affected by the time of diagnosis communication. Also, anticoagulation decreases short-term mortality for acute PE1014 that is quite high without appropriate treatment.1524 Thus, short-term mortality is another appropriate study end point in a relatively short follow-up period.

Our hospital introduced a policy in January 2006 that encourages radiologists to directly (ie, via face-to-face or telephone contact) notify responsible physicians of imaging findings that may threaten life and to document communication details in the report. Therefore, we have existing data regarding the communication details in each case of acute PE. The purpose of this study was to test the hypothesis that early direct communication of CT image findings is correlated with early treatment initiation and lower 30-day mortality.

Study Population and CTPA Imaging

The Partners institutional review board (#2010P001277) approved this Health Insurance Portability and Accountability Act-compliant study; informed consent was waived. Among the 1,177 CTPA examinations that were performed at our academic institution from February 2006 to March 2010 and were positive for acute PE, 796 patients met the inclusion criteria (Fig 1).

Figure Jump LinkFigure 1. Flow diagram for inclusion in the study. CTPA = CT pulmonary angiography; PE = pulmonary embolism.Grahic Jump Location

CTPA studies were performed by 16-, 64-, or 128-slice multidetector CT scanners with a standard protocol after IV administration of 75 to 100 mL iodinated contrast media at 3 to 4 mL/s. For each case, the right ventricle (RV) to left ventricle (LV) diameter ratio was measured on a reformatted four-chamber view25 as an indicator of RV dysfunction.26

Data Collection

The following data were obtained from the hospital electronic medical records: patient demographics (age, sex, race), admission to the ICU, hemodynamic status at initial presentation (hemodynamically unstable, defined as cardiac arrest or systolic BP < 90 mm Hg or a decrease in BP by > 40 mm Hg from baseline), comorbidities (cancer, congestive heart failure, coronary artery disease, atrial fibrillation, chronic lung disease, diabetes, hypertension, coagulopathies, chronic renal failure, or stroke), recent surgery (≤ 30 days), and therapy for PE (administration of therapeutic dose of anticoagulants, inferior vena cava [IVC] filter placement, thrombolysis, surgical or catheter thrombectomy). The time of therapy initiation was obtained from the electronic integrated clinical information system that recorded all ordering information at our institution. The proximal extension of the embolus (central, lobar, segmental, or subsegmental pulmonary artery), the time of CTPA study, and details regarding direct communication of findings were obtained from the radiology reports. The details of the communication are usually in or immediately following the Impression section of the report or as an addendum if the report was finalized before communication was complete. Documentation contains the name of the communicator, date and time reported, and name of recipient of the notification; for example, “Critical findings were communicated by Dr X to Dr Y at 5:00 pm on Wednesday, January 1, 2014.”

Study Outcomes

The first outcome was the time interval from CTPA acquisition to the initiation of PE-related therapy. Patients whose first PE-related treatment was a procedure (eg, IVC filter placement, thrombectomy, or catheter thrombolysis) were not included in this analysis, as the exact procedure time was difficult to determine. The second outcome was all-cause and PE-related death within 30 days of CTPA examination. Death was confirmed by the Social Security Death Index. To determine the cause of death for each case, three observers blinded to CT imaging interpretation or time to communication independently reviewed autopsy data, death certifications, electronic medical records, and physician notes from the referring physicians. The final decision was made by consensus; death was considered PE-related when (1) either the autopsy data, death certificate, or death report on the electronic medical record identified PE-related death, or (2) acute respiratory failure, cardiopulmonary arrest, or shock was the cause of death, in the absence of other cardiopulmonary diseases.

Study Exposure

The variable of interest was the time interval between CTPA acquisition and direct communication from the radiologist to the referring physician. This was categorized into early (≤ 1.5 h; n = 505) and late (> 1.5 h; n = 291) communications; the cutoff of 1.5 h was chosen based on the distribution of communication time interval with median and mean values of 1.2 and 2.0 h, respectively (Fig 2).

Figure Jump LinkFigure 2. Distribution of time interval from CTPA acquisition to direct communication, from CTPA acquisition to treatment initiation, and from communication to treatment initiation. Vertical lines are the median value (1.2 h), cutoff for dividing the early and late groups (1.5 h), and 98.5 percentile (15 h) for the time from CT to communication; median value (3.4 h) and 98.5 percentile (30 h) for the time from CT to treatment initiation; and the median value (2 h) for the time from communication to treatment initiation. See Figure 1 legend for expansion of abbreviation.Grahic Jump Location
Statistical Analysis

Univariate comparisons of clinical characteristics and outcomes between groups were performed using a χ2 test, unpaired Student t test, and Mann-Whitney U test, as appropriate. To analyze the relationship between the time to communication and the outcomes, we used the inverse probability weighting method by propensity scores for being in the late communication group to adjust for the potential confounders.27 For each patient, we calculated the propensity score for being in the late communication group as opposed to being in the early group, using a logistic regression model with the following variables: patient age, sex, race, ICU admission, hemodynamic instability, all comorbidities previously listed, enlarged RV on CTPA (RV/LV diameter ratio > 1.026) as a severity measure of PE, and time of CTPA (daytime vs nighttime [5:00 pm-7:00 am], weekdays vs weekends [from midnight Friday to midnight Sunday], and in earlier [2006, 2007] vs later years [2008-2010]). We then used propensity scores (PSs) to calculate the inverse probability weights for each patient to create a population where the distribution of the confounders were balanced in the early and the late communication groups: weight = (1−f(Late))/(1−PS) for the early communication group, and weight = f(late)/PS for the late communication group, where f(Late) is the proportion of those in the late group in the original population.28 The weighted Kaplan-Meier curves for the survival times and time to treatment initiation were drawn for each group. Finally, we ran a weighted Cox proportional hazards regression to estimate the hazard of each outcome for the late communication group relative to the early group and their 95% CIs. We also compared mortality among quartiles (n = 199 for each) based on time to treatment initiation both for all-cause and PE-related death.

We performed four sensitivity analyses: (1) including proximal extension of the clot as an indicator of PE severity instead of CT image-derived RV/LV diameter ratio; (2) excluding 1.5% of population with the longest communication time interval as outliers (Fig 2); (3) excluding 1.5% of population with the longest treatment initiation time interval as outliers (Fig 2); and (4) excluding observations with propensity scores either high or low enough to present only in either of the late or the early communication groups (nonoverlapping PS observations). Statistical analyses were performed using STATA version 10.1 (Stata Corp) and SAS version 9.3 software (SAS Institute Inc).

Clinical Characteristics

Patients in the early communication group (n = 505) had more severe PEs, indicated by the higher prevalence of central embolus and larger RV/LV diameter ratio (Table 1). Cancer, congestive heart failure, and chronic renal failure were more prevalent in the late communication group, and patients in this group were more likely to be admitted to the ICU. The CTPA studies in this group were more likely from the earlier years (2006-2007) compared with the early group.

Table Graphic Jump Location
Table 1 —Comparison of Clinical Characteristics

Continuous values expressed as mean ± SD. CTPA = CT pulmonary angiography; LV = left ventricle; PE = pulmonary embolism; RV = right ventricle.

a 

Early communication performed ≤ 1.5 h after CTPA.

b 

Late communication performed > 1.5 h after CTPA acquisition.

The distribution of propensity scores was similar between the early and late communication groups, with only 23 observations outside of the overlap. After inverse probability weighting, all 20 covariates were well balanced (P values ranged from .698 to .978).

Time From CTPA Acquisition to PE-Related Treatment Initiation

Table 2 summarizes the PE-related treatments. When warfarin was recorded as the first treatment order for PE, unfractionated heparin or low-molecular-weight heparin was also ordered almost simultaneously. Treatment characteristics were similar between groups, except for the higher rate of thrombolysis or thrombectomy in the early communication group. In 6.4% of patients (51 of 796), IVC filter placement (n = 50) or embolectomy (n = 1) was the first PE-related treatment; first outcome was not measured in this group. The proportion of these patients was not significantly different between early and late communication groups.

Table Graphic Jump Location
Table 2 —Comparison of PE-Related Treatments Between Early and Late Communication Groups

Data given as % unless otherwise indicated. IVC = inferior vena cava. See Table 1 legend for expansion of other abbreviation.

Figure 2 shows the distribution of the time from CTPA to treatment initiation, and the time from communication to treatment initiation. The treatment was initiated prior to the direct communication in 6.6% of patients (49 of 745), for whom all communication time intervals were > 60 min. In the univariate comparison, time to treatment was significantly longer in the late communication group (median, 3.3 vs 3.9 h for early vs late groups; P < .001) (Table 2). After controlling for confounders, compared with the early group, the late communication group was more likely to receive treatment later (adjusted hazard ratio [HR], 0.714; 95% CI, 0.610-0.836; P < .001) (Table 3). The weighted Kaplan-Meier curve is depicted in Figure 3A.

Table Graphic Jump Location
Table 3 —Results From Cox Regression Analyses Using Inverse Probability Weighting With Propensity Scores

SA = sensitivity analysis. See Table 1 legend for expansion of other abbreviations.

a 

Type of SA: SA 1 = a model including the proximal extension of the clot as an indicator of PE severity, instead of the CT scan-derived RV/LV diameter ratio; SA 2 = a model excluding 1.5% of population with the longest communication time interval as outliers; SA 3 = a model excluding 1.5% of population with the longest treatment initiation time interval as outliers; SA 4 = a model excluding observations with propensity scores in the nonoverlapping regions between the early and late communication groups.

b 

Hazard ratio for the late communication group, with the early group as a reference.

Figure Jump LinkFigure 3. Weighted Kaplan-Meier curves. A, Cumulative treatment initiation rate stratified by the early and late communications. B, Cumulative all-cause mortality rate stratified by the early and late communications. C, Cumulative pulmonary embolism-related mortality rate stratified by the early and late communications. Tx = pulmonary embolism-related treatment.Grahic Jump Location
Thirty-day Mortality

All-cause and PE-related 30-day mortality were 10.7% and 4.6%, respectively. The mortality rate was significantly higher in the late communication group (early vs late, 7.5% vs 16.1%, P < .001 for all-cause; 2.8% vs 7.9%, P = .001 for PE-related mortality) (Table 1). After controlling for confounders, Cox regression analysis showed significantly increased hazard of all-cause and PE-related mortality for the late group (adjusted HR, 1.813; 95% CI, 1.163-2.828; P = .009; and adjusted HR, 2.625; 95% CI, 1.362-5.059; P = .004, respectively) (Table 3). The weighted Kaplan-Meier curves are depicted in Figures 3B and 3C. Figure 4 shows 30-day PE-related mortality stratified by the time from CT scan to communication for every 30 min, demonstrating an increase of mortality after 1.5 h.

Figure Jump LinkFigure 4. PE-related 30-day mortality stratified by 30-min time increments to the time of communication. See Figure 1 legend for expansion of abbreviation.Grahic Jump Location
Sensitivity Analyses

Results from the four sensitivity analyses are presented in Table 3. They were qualitatively equal and quantitatively similar to those from the main analyses.

Mortality and Treatment Initiation Time

Figure 5 shows mortality rates after categorizing the population into quartiles based on the time interval between CTPA and treatment initiation. There was a significant trend for higher mortality with increasing time to treatment initiation for deaths unrelated to PE (all-cause minus PE-related deaths). Regarding PE-related death, there was a drop in the mortality in the fourth quartile (ie, the group with longest time to treatment).

Figure Jump LinkFigure 5. All-cause and PE-related 30-day mortality stratified by the quartile based on the time to treatment initiation interval. T-to-Com = median time interval from CT scan to communication; T-to-Tx = median time interval from CT scan to treatment. See Figure 1 legend for expansion of other abbreviation.Grahic Jump Location

Late (> 1.5 h from CTPA acquisition) communication of acute PE diagnosis from radiologists to referring physicians was significantly correlated with longer time to treatment initiation and higher 30-day mortality. Although we cannot draw conclusions on causality based solely on this observational study, our findings are, to our knowledge, the first evidence to support the general notion that earlier communication of imaging finding is beneficial to patient outcomes.

Although CTPA interpretation is relatively straightforward (ie, pulmonary artery filling defect), respiratory motion and poor contrast enhancement are challenging.29,30 In our cohort, a large majority of referring physicians started anticoagulation after the communication of the diagnosis from radiologists. We cannot tell whether referring physicians checked the images before the communication, but in either case, this finding provides evidence of the radiologist’s contribution to clinical decision making.

No published data address the factors influencing communication timing among radiologists. In our cohort, the early communication group had more severe cases of PE, in keeping with the premise that communication is more rapid when urgent care is indicated. On the other hand, comorbidities (eg, lung cancer) could lead to difficulty in image interpretation, resulting in longer communication times. These differences could confound the relationship between the exposure and outcomes, and, thus, require adjustments in the analysis. Inverse probability weighting31,32 with propensity scores3335 enabled adjustment for measured confounders without loss in stability of the analytic model in this cohort with a relatively small number of events. We also performed four different sensitivity analyses; two of these excluded potential outliers. Subjects with delayed communication or prolonged time to treatment could, in theory, have had special clinical scenarios that significantly influenced the estimation. Exclusion of outliers did not show a significant change in the estimates from the main analysis, and all four sensitivity analyses confirmed the robustness of the main result.

If expedited communication does, in fact, result in lowered mortality, the most possible intermediary pathway is through early initiation of treatment. This is indirectly supported by the higher HR of PE-related mortality compared with that of all-cause mortality. In a previous report, patients diagnosed with PE after admission had more inhospital adverse events (death, intubation, or circulatory shock) compared with those diagnosed in the ED.36 Hull et al37 did a pooled analysis of three anticoagulation trials, demonstrating that achieving a therapeutic activated partial thromboplastin time within 24 h after initiation of heparin reduced the risk of recurrent venous thromboembolism. Smith et al38 studied the relationship between survival and timing of anticoagulation, showing that early anticoagulation was associated with lower all-cause 30-day mortality. Our data (Fig 5) also suggest this association, except for the drop in PE-related mortality for the fourth quartile of the time to treatment. Patients in the fourth quartile may have had more severe comorbidities, suggested by the higher PE-unrelated mortality rate, that caused a delay in anticoagulation until its risk/benefit profile was assessed.

Although no standard communication time is established, literature recommends communication within 60 min for potentially immediately life-threatening findings.4,5 Given that this time frame is from the discovery of findings to the communication, our 1.5-h cutoff for early and late communications from the time of image acquisition seems appropriate. Current data suggesting that rapid communication appears rewarding, at least for patients with acute PE, will foster a future establishment of the worldwide standard for critical imaging results communication.

Besides being a single-center study that may limit the generalizability of our findings, we acknowledge several limitations to this study. First, we adjusted for prespecified potential confounders in the analysis, but we cannot exclude the possibility of remaining confounders, especially by unmeasureable factors. In theory, there is a limited number of possible factors affecting communication timing, because radiologists usually strive to deliver the CTPA findings as soon as possible. One possible example of an unmeasureable confounder is the workload or the busyness of the individual referring physician at the time of CTPA acquisition that affects both the initiation of treatment and how timely the communication from radiologists is established. While this factor, as well as other unmeasureable confounding factors, could, in part, contribute to an increased high HR, we do not expect these to completely explain the estimated HR of 2.6 for PE-related mortality. Second, we did not calculate clot burden3942 in this 796-patient cohort. Instead, a CT scan-derived RV/LV ratio and the location of the most proximal clot were included for adjustment of the severity of PE. Many studies do not show significant correlation between clot burden4248 and outcomes, while the RV/LV ratio has been reported as an independent predictor of short-term mortality.42 Third, we have excluded 121 patients for whom the exact time of direct communication was not available. This occurred when the radiologists neglected to document communication details. We, thus, assumed independence between the incomplete documentation and patient outcomes so that the exclusion of this subpopulation would not have significantly biased the estimation. In fact, the all-cause and PE-related 30-day mortalities were not significantly different between this population and the main population (13.2% vs 10.7%, P = .405; and 5.8% vs 4.6%, P = .586, respectively). Fourth, our institution does not have a system to record the exact time of medication administration; thus, we used the ordering system information to estimate the time of treatment initiation. There may have been a certain time difference between the order time and the actual administration time, but we believe that this time difference is, in theory, not dependent on the communication of imaging findings.

Late (> 1.5 h from CTPA acquisition) communication of acute PE diagnosis had a higher risk of delayed treatment initiation and higher 30-day mortality. In this retrospective observational study, these associations lend support to a potential contribution of early direct communication of CT image findings to clinical outcomes, and warrant a confirmation in larger studies.

Author contributions: Dr K. Kumamaru had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Dr K. Kumamaru: contributed to study concept and design; data acquisition, analysis, and interpretation; drafting the manuscript; approval of the final manuscript; administrative, technical, or material support; and study supervision and served as principal author.

Dr Hunsaker: contributed to study concept and design; administrative, technical, or material support; drafting the manuscript; and approval of the final manuscript.

Dr H. Kumamaru: contributed to data analysis and interpretation, drafting the manuscript, and approval of the final manuscript and provided statistical expertise.

Dr George: contributed to data acquisition, drafting the manuscript, and approval of the final manuscript.

Dr Bedayat: contributed to data acquisition, drafting the manuscript, and approval of the final manuscript.

Dr Rybicki: contributed to study concept and design; administrative, technical, or material support; study supervision; drafting the manuscript; and approval of the final manuscript.

Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Role of sponsors: The Japan Society for the Promotion of Science was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

CTPA

CT pulmonary angiography

IVC

inferior vena cava

LV

left ventricle

PE

pulmonary embolism

PS

propensity score

RV

right ventricle

Kushner DC, Lucey LL; American College of Radiology. Diagnostic radiology reporting and communication: the ACR guideline. J Am Coll Radiol. 2005;2(1):15-21. [CrossRef] [PubMed]
 
Lucey LL, Kushner DC; American College of Radiology. The ACR guideline on communication: to be or not to be, that is the question. J Am Coll Radiol. 2010;7(2):109-114. [CrossRef] [PubMed]
 
Solomon R. The role of osmolality in the incidence of contrast-induced nephropathy: a systematic review of angiographic contrast media in high risk patients. Kidney Int. 2005;68(5):2256-2263. [CrossRef] [PubMed]
 
Anthony SG, Prevedello LM, Damiano MM, et al. Impact of a 4-year quality improvement initiative to improve communication of critical imaging test results. Radiology. 2011;259(3):802-807. [CrossRef] [PubMed]
 
Khorasani R. Optimizing communication of critical test results. J Am Coll Radiol. 2009;6(10):721-723. [CrossRef] [PubMed]
 
Stein PD, Fowler SE, Goodman LR, et al; PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med. 2006;354(22):2317-2327. [CrossRef] [PubMed]
 
Hunsaker AR, Lu MT, Goldhaber SZ, Rybicki FJ. Imaging in acute pulmonary embolism with special clinical scenarios. Circ Cardiovasc Imaging. 2010;3(4):491-500. [CrossRef] [PubMed]
 
van Belle A, Büller HR, Huisman MV, et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA. 2006;295(2):172-179. [CrossRef] [PubMed]
 
Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2_suppl):e419S-494S
 
Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism. A controlled trial. Lancet. 1960;1(7138):1309-1312. [CrossRef] [PubMed]
 
Kernohan RJ, Todd C. Heparin therapy in thromboembolic disease. Lancet. 1966;1(7438):621-623. [CrossRef] [PubMed]
 
Alpert JS, Smith R, Carlson J, Ockene IS, Dexter L, Dalen JE. Mortality in patients treated for pulmonary embolism. JAMA. 1976;236(13):1477-1480. [CrossRef] [PubMed]
 
Kanis JA. Heparin in the treatment of pulmonary thromboembolism. Thromb Diath Haemorrh. 1974;32(2-3):519-527. [PubMed]
 
Raschke RA, Reilly BM, Guidry JR, Fontana JR, Srinivas S. The weight-based heparin dosing nomogram compared with a “standard care” nomogram. A randomized controlled trial. Ann Intern Med. 1993;119(9):874-881. [CrossRef] [PubMed]
 
Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979-1998: an analysis using multiple-cause mortality data. Arch Intern Med. 2003;163(14):1711-1717. [CrossRef] [PubMed]
 
Dismuke SE, Wagner EH. Pulmonary embolism as a cause of death. The changing mortality in hospitalized patients. JAMA. 1986;255(15):2039-2042. [CrossRef] [PubMed]
 
Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovasc Dis. 1975;17(4):259-270. [CrossRef] [PubMed]
 
Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med. 1991;151(5):933-938. [CrossRef] [PubMed]
 
Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet. 1999;353(9162):1386-1389. [CrossRef] [PubMed]
 
Goldhaber SZ, Morpurgo M; Report of the WHO/International Society and Federation of Cardiology Task Force. Diagnosis, treatment, and prevention of pulmonary embolism. JAMA. 1992;268(13):1727-1733. [CrossRef] [PubMed]
 
Barritt DW, Jordan SC. Clinical features of pulmonary embolism. Lancet. 1961;1(7180):729-732. [CrossRef] [PubMed]
 
Carson JL, Kelley MA, Duff A, et al. The clinical course of pulmonary embolism. N Engl J Med. 1992;326(19):1240-1245. [CrossRef] [PubMed]
 
Nijkeuter M, Söhne M, Tick LW, et al; Christopher Study Investigators. The natural course of hemodynamically stable pulmonary embolism: Clinical outcome and risk factors in a large prospective cohort study. Chest. 2007;131(2):517-523. [CrossRef] [PubMed]
 
Moser KM. Venous thromboembolism. Am Rev Respir Dis. 1990;141(1):235-249. [CrossRef] [PubMed]
 
Lu MT, Ersoy H, Whitmore AG, Lipton MJ, Rybicki FJ. Reformatted four-chamber and short-axis views of the heart using thin section (</=2 mm) MDCT images. Acad Radiol. 2007;14(9):1108-1112. [CrossRef] [PubMed]
 
Kumamaru KK, Lu MT, Ghaderi Niri S, Hunsaker AR. Right ventricular enlargement in acute pulmonary embolism derived from CT pulmonary angiography. Int J Cardiovasc Imaging. 2013;29(3):705-708. [CrossRef] [PubMed]
 
Austin PC. The performance of different propensity score methods for estimating marginal hazard ratios. Stat Med. 2013;32(16):2837-2849. [CrossRef] [PubMed]
 
Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11(5):550-560. [CrossRef] [PubMed]
 
Jones SE, Wittram C. The indeterminate CT pulmonary angiogram: imaging characteristics and patient clinical outcome. Radiology. 2005;237(1):329-337. [CrossRef] [PubMed]
 
Wittram C, Maher MM, Yoo AJ, Kalra MK, Shepard JA, McLoud TC. CT angiography of pulmonary embolism: diagnostic criteria and causes of misdiagnosis. Radiographics. 2004;24(5):1219-1238. [CrossRef] [PubMed]
 
Lunceford JK, Davidian M. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med. 2004;23(19):2937-2960. [CrossRef] [PubMed]
 
Rosenbaum P. Model-based direct adjustment. J Am Stat Assoc. 1987;82(398):387-394. [CrossRef]
 
Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. [CrossRef] [PubMed]
 
Rosenbaum P, Rubin D. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70(1):41-55. [CrossRef]
 
Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol. 1999;150(4):327-333. [CrossRef] [PubMed]
 
Kline JA, Hernandez-Nino J, Jones AE, Rose GA, Norton HJ, Camargo CA Jr. Prospective study of the clinical features and outcomes of emergency department patients with delayed diagnosis of pulmonary embolism. Acad Emerg Med. 2007;14(7):592-598. [PubMed]
 
Hull RD, Raskob GE, Brant RF, Pineo GF, Valentine KA. Relation between the time to achieve the lower limit of the APTT therapeutic range and recurrent venous thromboembolism during heparin treatment for deep vein thrombosis. Arch Intern Med. 1997;157(22):2562-2568. [CrossRef] [PubMed]
 
Smith SB, Geske JB, Maguire JM, Zane NA, Carter RE, Morgenthaler TI. Early anticoagulation is associated with reduced mortality for acute pulmonary embolism. Chest. 2010;137(6):1382-1390. [CrossRef] [PubMed]
 
Qanadli SD, El Hajjam M, Vieillard-Baron A, et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR Am J Roentgenol. 2001;176(6):1415-1420. [CrossRef] [PubMed]
 
Mastora I, Remy-Jardin M, Masson P, et al. Severity of acute pulmonary embolism: evaluation of a new spiral CT angiographic score in correlation with echocardiographic data. Eur Radiol. 2003;13(1):29-35. [PubMed]
 
Furlan A, Patil A, Park B, Chang CC, Roberts MS, Bae KT. Accuracy and reproducibility of blood clot burden quantification with pulmonary CT angiography. AJR Am J Roentgenol. 2011;196(3):516-523. [CrossRef] [PubMed]
 
Furlan A, Aghayev A, Chang CC, et al. Short-term mortality in acute pulmonary embolism: clot burden and signs of right heart dysfunction at CT pulmonary angiography. Radiology. 2012;265(1):283-293. [CrossRef] [PubMed]
 
Ghaye B, Ghuysen A, Willems V, et al. Severe pulmonary embolism:pulmonary artery clot load scores and cardiovascular parameters as predictors of mortality. Radiology. 2006;239(3):884-891. [CrossRef] [PubMed]
 
Ghuysen A, Ghaye B, Willems V, et al. Computed tomographic pulmonary angiography and prognostic significance in patients with acute pulmonary embolism. Thorax. 2005;60(11):956-961. [CrossRef] [PubMed]
 
Araoz PA, Gotway MB, Harrington JR, Harmsen WS, Mandrekar JN. Pulmonary embolism: prognostic CT findings. Radiology. 2007;242(3):889-897. [CrossRef] [PubMed]
 
Araoz PA, Gotway MB, Trowbridge RL, et al. Helical CT pulmonary angiography predictors of in-hospital morbidity and mortality in patients with acute pulmonary embolism. J Thorac Imaging. 2003;18(4):207-216. [CrossRef] [PubMed]
 
Pech M, Wieners G, Dul P, et al. Computed tomography pulmonary embolism index for the assessment of survival in patients with pulmonary embolism. Eur Radiol. 2007;17(8):1954-1959. [CrossRef] [PubMed]
 
Vedovati MC, Becattini C, Agnelli G, et al. Multidetector CT scan for acute pulmonary embolism: embolic burden and clinical outcome. Chest. 2012;142(6):1417-1424. [CrossRef] [PubMed]
 

Figures

Figure Jump LinkFigure 1. Flow diagram for inclusion in the study. CTPA = CT pulmonary angiography; PE = pulmonary embolism.Grahic Jump Location
Figure Jump LinkFigure 2. Distribution of time interval from CTPA acquisition to direct communication, from CTPA acquisition to treatment initiation, and from communication to treatment initiation. Vertical lines are the median value (1.2 h), cutoff for dividing the early and late groups (1.5 h), and 98.5 percentile (15 h) for the time from CT to communication; median value (3.4 h) and 98.5 percentile (30 h) for the time from CT to treatment initiation; and the median value (2 h) for the time from communication to treatment initiation. See Figure 1 legend for expansion of abbreviation.Grahic Jump Location
Figure Jump LinkFigure 3. Weighted Kaplan-Meier curves. A, Cumulative treatment initiation rate stratified by the early and late communications. B, Cumulative all-cause mortality rate stratified by the early and late communications. C, Cumulative pulmonary embolism-related mortality rate stratified by the early and late communications. Tx = pulmonary embolism-related treatment.Grahic Jump Location
Figure Jump LinkFigure 4. PE-related 30-day mortality stratified by 30-min time increments to the time of communication. See Figure 1 legend for expansion of abbreviation.Grahic Jump Location
Figure Jump LinkFigure 5. All-cause and PE-related 30-day mortality stratified by the quartile based on the time to treatment initiation interval. T-to-Com = median time interval from CT scan to communication; T-to-Tx = median time interval from CT scan to treatment. See Figure 1 legend for expansion of other abbreviation.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1 —Comparison of Clinical Characteristics

Continuous values expressed as mean ± SD. CTPA = CT pulmonary angiography; LV = left ventricle; PE = pulmonary embolism; RV = right ventricle.

a 

Early communication performed ≤ 1.5 h after CTPA.

b 

Late communication performed > 1.5 h after CTPA acquisition.

Table Graphic Jump Location
Table 2 —Comparison of PE-Related Treatments Between Early and Late Communication Groups

Data given as % unless otherwise indicated. IVC = inferior vena cava. See Table 1 legend for expansion of other abbreviation.

Table Graphic Jump Location
Table 3 —Results From Cox Regression Analyses Using Inverse Probability Weighting With Propensity Scores

SA = sensitivity analysis. See Table 1 legend for expansion of other abbreviations.

a 

Type of SA: SA 1 = a model including the proximal extension of the clot as an indicator of PE severity, instead of the CT scan-derived RV/LV diameter ratio; SA 2 = a model excluding 1.5% of population with the longest communication time interval as outliers; SA 3 = a model excluding 1.5% of population with the longest treatment initiation time interval as outliers; SA 4 = a model excluding observations with propensity scores in the nonoverlapping regions between the early and late communication groups.

b 

Hazard ratio for the late communication group, with the early group as a reference.

References

Kushner DC, Lucey LL; American College of Radiology. Diagnostic radiology reporting and communication: the ACR guideline. J Am Coll Radiol. 2005;2(1):15-21. [CrossRef] [PubMed]
 
Lucey LL, Kushner DC; American College of Radiology. The ACR guideline on communication: to be or not to be, that is the question. J Am Coll Radiol. 2010;7(2):109-114. [CrossRef] [PubMed]
 
Solomon R. The role of osmolality in the incidence of contrast-induced nephropathy: a systematic review of angiographic contrast media in high risk patients. Kidney Int. 2005;68(5):2256-2263. [CrossRef] [PubMed]
 
Anthony SG, Prevedello LM, Damiano MM, et al. Impact of a 4-year quality improvement initiative to improve communication of critical imaging test results. Radiology. 2011;259(3):802-807. [CrossRef] [PubMed]
 
Khorasani R. Optimizing communication of critical test results. J Am Coll Radiol. 2009;6(10):721-723. [CrossRef] [PubMed]
 
Stein PD, Fowler SE, Goodman LR, et al; PIOPED II Investigators. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med. 2006;354(22):2317-2327. [CrossRef] [PubMed]
 
Hunsaker AR, Lu MT, Goldhaber SZ, Rybicki FJ. Imaging in acute pulmonary embolism with special clinical scenarios. Circ Cardiovasc Imaging. 2010;3(4):491-500. [CrossRef] [PubMed]
 
van Belle A, Büller HR, Huisman MV, et al; Christopher Study Investigators. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA. 2006;295(2):172-179. [CrossRef] [PubMed]
 
Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2_suppl):e419S-494S
 
Barritt DW, Jordan SC. Anticoagulant drugs in the treatment of pulmonary embolism. A controlled trial. Lancet. 1960;1(7138):1309-1312. [CrossRef] [PubMed]
 
Kernohan RJ, Todd C. Heparin therapy in thromboembolic disease. Lancet. 1966;1(7438):621-623. [CrossRef] [PubMed]
 
Alpert JS, Smith R, Carlson J, Ockene IS, Dexter L, Dalen JE. Mortality in patients treated for pulmonary embolism. JAMA. 1976;236(13):1477-1480. [CrossRef] [PubMed]
 
Kanis JA. Heparin in the treatment of pulmonary thromboembolism. Thromb Diath Haemorrh. 1974;32(2-3):519-527. [PubMed]
 
Raschke RA, Reilly BM, Guidry JR, Fontana JR, Srinivas S. The weight-based heparin dosing nomogram compared with a “standard care” nomogram. A randomized controlled trial. Ann Intern Med. 1993;119(9):874-881. [CrossRef] [PubMed]
 
Horlander KT, Mannino DM, Leeper KV. Pulmonary embolism mortality in the United States, 1979-1998: an analysis using multiple-cause mortality data. Arch Intern Med. 2003;163(14):1711-1717. [CrossRef] [PubMed]
 
Dismuke SE, Wagner EH. Pulmonary embolism as a cause of death. The changing mortality in hospitalized patients. JAMA. 1986;255(15):2039-2042. [CrossRef] [PubMed]
 
Dalen JE, Alpert JS. Natural history of pulmonary embolism. Prog Cardiovasc Dis. 1975;17(4):259-270. [CrossRef] [PubMed]
 
Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. The Worcester DVT Study. Arch Intern Med. 1991;151(5):933-938. [CrossRef] [PubMed]
 
Goldhaber SZ, Visani L, De Rosa M. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER). Lancet. 1999;353(9162):1386-1389. [CrossRef] [PubMed]
 
Goldhaber SZ, Morpurgo M; Report of the WHO/International Society and Federation of Cardiology Task Force. Diagnosis, treatment, and prevention of pulmonary embolism. JAMA. 1992;268(13):1727-1733. [CrossRef] [PubMed]
 
Barritt DW, Jordan SC. Clinical features of pulmonary embolism. Lancet. 1961;1(7180):729-732. [CrossRef] [PubMed]
 
Carson JL, Kelley MA, Duff A, et al. The clinical course of pulmonary embolism. N Engl J Med. 1992;326(19):1240-1245. [CrossRef] [PubMed]
 
Nijkeuter M, Söhne M, Tick LW, et al; Christopher Study Investigators. The natural course of hemodynamically stable pulmonary embolism: Clinical outcome and risk factors in a large prospective cohort study. Chest. 2007;131(2):517-523. [CrossRef] [PubMed]
 
Moser KM. Venous thromboembolism. Am Rev Respir Dis. 1990;141(1):235-249. [CrossRef] [PubMed]
 
Lu MT, Ersoy H, Whitmore AG, Lipton MJ, Rybicki FJ. Reformatted four-chamber and short-axis views of the heart using thin section (</=2 mm) MDCT images. Acad Radiol. 2007;14(9):1108-1112. [CrossRef] [PubMed]
 
Kumamaru KK, Lu MT, Ghaderi Niri S, Hunsaker AR. Right ventricular enlargement in acute pulmonary embolism derived from CT pulmonary angiography. Int J Cardiovasc Imaging. 2013;29(3):705-708. [CrossRef] [PubMed]
 
Austin PC. The performance of different propensity score methods for estimating marginal hazard ratios. Stat Med. 2013;32(16):2837-2849. [CrossRef] [PubMed]
 
Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11(5):550-560. [CrossRef] [PubMed]
 
Jones SE, Wittram C. The indeterminate CT pulmonary angiogram: imaging characteristics and patient clinical outcome. Radiology. 2005;237(1):329-337. [CrossRef] [PubMed]
 
Wittram C, Maher MM, Yoo AJ, Kalra MK, Shepard JA, McLoud TC. CT angiography of pulmonary embolism: diagnostic criteria and causes of misdiagnosis. Radiographics. 2004;24(5):1219-1238. [CrossRef] [PubMed]
 
Lunceford JK, Davidian M. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Stat Med. 2004;23(19):2937-2960. [CrossRef] [PubMed]
 
Rosenbaum P. Model-based direct adjustment. J Am Stat Assoc. 1987;82(398):387-394. [CrossRef]
 
Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424. [CrossRef] [PubMed]
 
Rosenbaum P, Rubin D. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70(1):41-55. [CrossRef]
 
Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol. 1999;150(4):327-333. [CrossRef] [PubMed]
 
Kline JA, Hernandez-Nino J, Jones AE, Rose GA, Norton HJ, Camargo CA Jr. Prospective study of the clinical features and outcomes of emergency department patients with delayed diagnosis of pulmonary embolism. Acad Emerg Med. 2007;14(7):592-598. [PubMed]
 
Hull RD, Raskob GE, Brant RF, Pineo GF, Valentine KA. Relation between the time to achieve the lower limit of the APTT therapeutic range and recurrent venous thromboembolism during heparin treatment for deep vein thrombosis. Arch Intern Med. 1997;157(22):2562-2568. [CrossRef] [PubMed]
 
Smith SB, Geske JB, Maguire JM, Zane NA, Carter RE, Morgenthaler TI. Early anticoagulation is associated with reduced mortality for acute pulmonary embolism. Chest. 2010;137(6):1382-1390. [CrossRef] [PubMed]
 
Qanadli SD, El Hajjam M, Vieillard-Baron A, et al. New CT index to quantify arterial obstruction in pulmonary embolism: comparison with angiographic index and echocardiography. AJR Am J Roentgenol. 2001;176(6):1415-1420. [CrossRef] [PubMed]
 
Mastora I, Remy-Jardin M, Masson P, et al. Severity of acute pulmonary embolism: evaluation of a new spiral CT angiographic score in correlation with echocardiographic data. Eur Radiol. 2003;13(1):29-35. [PubMed]
 
Furlan A, Patil A, Park B, Chang CC, Roberts MS, Bae KT. Accuracy and reproducibility of blood clot burden quantification with pulmonary CT angiography. AJR Am J Roentgenol. 2011;196(3):516-523. [CrossRef] [PubMed]
 
Furlan A, Aghayev A, Chang CC, et al. Short-term mortality in acute pulmonary embolism: clot burden and signs of right heart dysfunction at CT pulmonary angiography. Radiology. 2012;265(1):283-293. [CrossRef] [PubMed]
 
Ghaye B, Ghuysen A, Willems V, et al. Severe pulmonary embolism:pulmonary artery clot load scores and cardiovascular parameters as predictors of mortality. Radiology. 2006;239(3):884-891. [CrossRef] [PubMed]
 
Ghuysen A, Ghaye B, Willems V, et al. Computed tomographic pulmonary angiography and prognostic significance in patients with acute pulmonary embolism. Thorax. 2005;60(11):956-961. [CrossRef] [PubMed]
 
Araoz PA, Gotway MB, Harrington JR, Harmsen WS, Mandrekar JN. Pulmonary embolism: prognostic CT findings. Radiology. 2007;242(3):889-897. [CrossRef] [PubMed]
 
Araoz PA, Gotway MB, Trowbridge RL, et al. Helical CT pulmonary angiography predictors of in-hospital morbidity and mortality in patients with acute pulmonary embolism. J Thorac Imaging. 2003;18(4):207-216. [CrossRef] [PubMed]
 
Pech M, Wieners G, Dul P, et al. Computed tomography pulmonary embolism index for the assessment of survival in patients with pulmonary embolism. Eur Radiol. 2007;17(8):1954-1959. [CrossRef] [PubMed]
 
Vedovati MC, Becattini C, Agnelli G, et al. Multidetector CT scan for acute pulmonary embolism: embolic burden and clinical outcome. Chest. 2012;142(6):1417-1424. [CrossRef] [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543