Incidence and Prevention of Postoperative Venous Thromboembolism: Are They Meaningful Quality Indicators in Japanese Health Care Settings?
Venous thromboembolism (VTE) epidemiology varies widely across surgical procedures. At present, there are few epidemiologic reports regarding VTE in Japan. Japanese VTE prophylaxis guidelines recommend a risk-based approach based on previous epidemiologic statistics. VTE includes deep vein thrombosis (DVT) and pulmonary embolism (PE). PE prevention is the main goal, although the relation between PE and DVT is still controversial.
We collected administrative data for 1,016,496 surgical patients from 260 hospitals. We analyzed DVT and PE incidence and selected two subgroups for further analysis: gastroenterologic surgery and specific orthopedic surgery (high-frequency group).
Overall DVT incidence was 1947 (0.19%); and the PE incidence was 538 (0.05%). DVT case fatality rate was 3.44% (67/1947); that for PE was 22.86% (123/538). Both overall and subgroup incidences were comparable to those in previous reports. Subgroup analyses in the high-frequency group did not show a relation between DVT and PE. VTE prophylaxis did not show a relation between DVT and PE despite 99% adherence.
Our results are consistent with established data regarding DVT and PE incidence. Administrative data available in Japan provides a powerful epidemiologic tool to characterize rare diseases such as DVT and PE. DVT is not a suitable quality indicator in Japan. However, PE is too rare to be considered a rate-based outcome indicator, and VTE prophylaxis is too widely applied to be used as a process indicator. VTE measurement is not a useful quality indicator in Japan to compare hospitals but provides a longitudinal self-survey.
Venous thromboembolism (VTE) is known as a major complication associated with surgery and comprises deep vein thrombosis (DVT) and pulmonary embolism (PE). Several guidelines for VTE prophylaxis have been published and adapted for clinical practice [1, 2, 3, 4, 5]. VTE occurs relatively less frequently in Japan than in other countries, and there have been few reports of its epidemiology .
The Japanese medical insurance system introduced the diagnosis–procedure combination (DPC)-based reimbursement system in 2003. To participate in this system, each hospital is required to collect clinical information as well as an applied reimbursement record. We first analyzed the epidemiology of VTE associated with surgery in Japan using the DPC database. As the epidemiology of VTE is known to vary across surgical procedures, we also selected two subgroups and analyzed them in detail.
Since guidelines were revised in 2004 , health care providers have become more interested in VTE quality indicators at the hospital level. In this study, we refer to quality indicators as “measures of health care quality that make use of readily available hospital inpatient administrative data” . Although the guideline’s most important goal is to prevent PE by thromboprophylaxis, the relation between DVT and PE is controversial, especially in patients undergoing hip or knee replacement [1, 2, 7]. In this study, we analyzed the relation between DVT and PE by comparing their incidence and that of VTE prophylaxis among hospitals. Furthermore, we considered whether a VTE quality indicator would have meaningful application in Japan.
The Kyoto University Graduate School of Medicine’s institutional review board approved the study protocol. Data were extracted from the Quality Indicator/Improvement Project (QIP) database. QIP collects DPC data from hospitals with the purpose of analyzing health care processes, patient outcomes, and disease management to provide feedback to the participating hospitals. Hospitals in the QIP have voluntarily joined the project and represent a variety of public, private, teaching, and nonteaching hospitals with different proportions of cases and specialties. A total of 260 hospitals were included in these analyses.
The DPC data are created for each patient per hospitalization. In addition to the hospital and patient identifier and the DPC codes, the data also include items such as the principal diagnosis (one field), the major diagnosis (one field), primary diagnosis (one field), secondary diagnosis (one field), co-morbidities and complications (eight fields), and surgery information.
By combining major diagnostic data and surgery information, it was possible to determine the specific surgery for each case. We used approximately 2 million consecutive records of patients discharged from 260 hospitals between April 2008 and March 2010. From the database, we selected patients who had one or more surgeries and who were without complications associated with VTE at the time of admission. Surgical cases were selected using K-codes, which are Japanese surgical codes used for insurance reimbursement, excluding blood infusion only and surgery for embolisms (see Electronic Supplementary Material 1). VTE was identified using ICD-10 codes including DVT (I801, I802, I809, I82*, O223, O229, O871, O878, and O879) and PE (I26*, O882, and O888). A total of 1,016,496 cases were included in the analyses.
We selected two subgroups for detailed analyses. The first group was a gastroenterologic surgery group, which is commonly conducted in Japan. This group included patients who had undergone gastroenterologic surgery, making possible comparisons with open and laparoscopic procedures for treating benign and malignant tumors. Gastroenterologic surgeries were identified by surgery information using K-codes including those for open surgery for nonmalignancies, laparoscopic surgery for nonmalignancies, open surgery for malignancies, and laparoscopic surgery for malignancies (see Electronic Supplementary Material 1).
The second group consisted of patients who had undergone specific orthopedic surgery, representing a high-frequency group for VTE incidence. We placed six of the largest number of incidence categories into this group using combinations of ICD-10 codes (major diagnosis codes) and K-codes (see Electronic Supplementary Material 1), including hip arthroplasty, knee arthroplasty, open reduction for femoral fractures, arthroplasty for femoral fractures, open reduction for lower leg fractures, and re-arthroplasty. In particular, total hip or knee arthroplasty are known high-risk factors [1, 2, 3, 4, 5]. We selected this high-frequency group for further analysis because of its relative homogeneity in procedures, thereby providing a larger number of incidents for analysis.
Overall incidence of thrombotic complications in patients who underwent any surgery
Others or none
General + epidural
The incidence of PE observed here was close to the incidence reported in another study, which used a denominator of patients administered to by an anesthesia specialist . Women were likely to develop both DVT and PE, and aging seems to be an increasingly important factor. DVT and PE were more frequently seen with general anesthesia than with epidural anesthesia alone, and some cases occurred in patients undergoing surgery even under local anesthesia.
Overall mortality for patients who underwent any surgery
Deep vein thrombosis and pulmonary embolism incidence in the gastroenterologic surgery group
Open surgery for nonmalignancy
Laparoscopic surgery for nonmalignancy
Open surgery for malignancy
Laparoscopic surgery for malignancy
Others or none
General + epidural
Mortality for the gastroenterologic surgery group
The difference in the incidence between the sexes in this subgroup was smaller than that for all cases. Although laparoscopic surgery used to be considered a risk factor because of artificial pneumoperitoneum , it caused VTE less often than open surgery here. This tendency is consistent with the estimate published by the American College of Chest Physicians (ACCP) . Surgery to treat malignancies caused more VTE than surgery to treat nonmalignancies. Aging seemed to be an increasingly important factor in this group.
Incidence of deep vein thrombosis and pulmonary embolism in the high-frequency group
Open reduction for femoral fractures
Arthroplasty for femoral fractures
Open reduction for lower leg fractures
Others or none
General + epidural
Mortality of the high-frequency group
The incidence of VTE and PE associated with hip replacement were, respectively, 2.85% and 0.24%. VTE and PE incidences for knee replacement were 2.28% and 0.20%, respectively. The incidence of VTE that we determined here was lower than that stated in the guidelines, whereas that of PE was similar .
Both DVT and PE proportions were higher than in the overall analysis, and they were more frequent in women. Aging did not appear to be an enhancing factor. Although PE was associated with a relatively high mortality, it was not as high in this group as in the overall data.
According to the guidelines, this group is considered at moderate to high risk for VTE ; therefore, we assumed that prophylactic treatment would have been recommended to all patients either in the form of anticoagulation or mechanical methods. We combined procedure data for each case, which were identified by drugs administered and by the embolic protection management fee. If the management fee is calculated without considering anticoagulation drugs, we assume the patient received a mechanical thromboprophylaxis treatment.
We studied the epidemiology of VTE because few data are available on this critically important surgical complication, especially in Japan. One reason for this is that VTE is relatively rare among Asians , and the small number of cases make epidemiologic studies problematic. Therefore, we took advantage of Japan’s DPC data system, which collects data in the same format from many hospitals and is thus useful for analyzing rare diseases. Because the calculated incidences were consistent with published data, we judged the DPC data as sufficiently reliable for analysis. Moreover, because our subgroup epidemiologic analysis method was reliable, future researchers will be able to analyze any other subgroups in which they are interested.
Because VTE prophylaxis guidelines are already published, quality indicators for VTE are recommended. Quality indicators help measure health care quality by using readily available hospital inpatient administrative data , which enable us to measure objectively the numeric barometers of medical processes, outcomes, and economic efficiency. These indicators can be surveyed by each hospital or by feedback to the medical providers such as the QIP—administered by our department. These “Quality Indicators can be used to highlight potential quality concerns, identify areas that need further study and investigation, and track changes over time”  so that each hospital can be aware of its status in the context of other hospitals and eventually make necessary improvements on their performance and quality of health care. For example, the Joint Commission suggests six measures related to VTE , and it therefore appears potentially possible to assess VTE incidence as a quality measure. However, this assessment is justified only when there is direct relation between DVT and PE because PE, but not DVT, is associated with high mortality and our main goal is to prevent PE. This relation between PE and DVT is still controversial. Whereas the ACCP believes that the relation exists, the American Academy of Orthopaedic Surgeons doubts its existence, particularly in the context of hip and knee arthroplasty surgery. In our subgroup analysis, no relation between DVT and PE incidence was detected. If PE was assumed to be a consequence of DVT at a finite probability, the PE incidence would have correlated with that of DVT.
There are some possible explanations for differences in DVT incidence among hospitals. For example, hospitals that have not adopted routine examinations to detect DVT may tend to underdiagnose the condition compared to those that have adopted routine examinations. Another possibility is that because DVT shows varying degrees of severity and low mortality, physicians may have used different diagnostic criteria. Furthermore, DVT may consist of potentially different subgroups that are accompanied by PE either frequently or infrequently.
Ideal indicators must prospectively result in better outcomes and should be specific and sensitive . According to these criteria, DVT cannot be included as an indicator; instead, PE should be considered alone.
However, the incidence of PE is very low in Japan, as illustrated by our analysis of the high-frequency group, which indicated that the incidence averaged 0.2%. Furthermore, the average number of patients in this group was 204, and the highest incidence in a hospital was six cases in 2 years. Thus, the PE incidence appears inconsistent across time and is too small to be considered a rate-based indicator. Moreover, risk adjustments are difficult for conditions with low incidences.
In contrast, 99% of patients with DVT and PE in this group underwent prophylactic interventions (88–100%, by hospitals). This finding suggests that PE can develop in patients who receive such therapy and those who do not; it also shows that hospitals are highly aware of VTE prophylaxis. Considering these contraindications, the true proportion of hospitals adopting VTE prophylaxis should be higher. Thus, a process indicator is not useful in Japan.
The incidence of VTE in the group given an antithrombotic agent was higher than the groups that received either no treatment or mechanical intervention alone. According to the Japanese guidelines, mechanical methods are recommended by default. Anticoagulation is not routinely recommended and is used on a case-by-case basis, especially in high-risk patients. However, high-risk patients are naturally at an increased risk of VTE; and despite treatment with anticoagulation methods, they may be more prone to thrombotic incidents.
The incidence of PE is so low in Japan that it may seem to serve as a candidate sentinel indicator. Such sentinel indicators refer to incidents that require further assessment to search for any poor practices or procedures . However, PE does not always occur accidentally, thus rendering PE events not suitable for assessing hospital quality. Consideration of each PE episode is important, and the etiology of PE should be researched thoroughly by Japanese clinicians. Using DVT and PE incidences as indicators of functioning of hospitals, however, can lead to incorrect assessments.
The strength of this study is the use of the large data set obtained from the Japanese administrative database under the DPC system, which allows greater representation of acute care hospitals in general. One limitation of this database is that it consists only of form-filling data input, which may imply a lack of sufficient clinical data available for analysis. However, it is because these forms are uniformly formatted that we can collect, handle, and analyze this database with relative ease. Another limitation is that the database in this study did not involve random sampling, although the data were obtained from various types of hospitals in different regions in Japan. The use of such widely varied sources supports the assumption that our database is not intrinsically biased and includes numerous variations due to hospital types, size, practice patterns, and geographic variations. Our results were close to those from another PE incidence data analysis from the Annual Study of Japanese Society of Anesthesiologists, which was based on data from more than 1000 hospitals . Although our results are similar to those in previous studies, the novelty of our study lies in the use of the different data source, whereas previous studies have mostly used relatively small samples, data from questionnaires, meta-analyses of questionnaires and systematic reviews [2, 4], or data limited to anesthesiologists . This study includes various clinical cases, not just those under an anesthesiologist’s care, as well as an analyses of both DVT and PE incidences. Another advantage of this research lies in the use of a new and simpler methodology derived from an existing database that is consistently updated to include data from recent admissions, thereby lending itself to application in large-scale multiinstitutional analysis, which has not been seen previously in Japan. The Japanese DPC database is collected every year, and DPC database analysis may increase in the future. As such, this is a pioneering study in VTE using the DPC database, demonstrating not only that the database is an efficient one for such studies but also that we have to be cautious in the use of such indicators as quality measures.
We have shown that Japanese hospital administrative data support reliable epidemiologic studies of VTE. We also illustrated this database’s usefulness for analyzing diseases with low incidence. While VTE incidence and preventive procedures are being used as quality indicators in several other countries, the low PE incidence in Japan precludes its use as a health care outcome indicator in this country. As VTE prophylaxis appears to be sufficiently applied in Japan, it lacks usefulness as a process indicator. We conclude that rate-based indicators of VET and PE would not be useful for evaluating health care in Japanese hospitals.
This study was supported in part by a Health Sciences Research Grant from the Ministry of Health, Labor, and Welfare of Japan and a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science.
Conflicts of interest
The authors declare that there are no potential or real conflicts of interest.
- 2.American Academy of Orthopaedic Surgeons Board of Directors (2007) American Academy of Orthopaedic Surgeons Clinical Guideline on Prevention of Pulmonary Embolism in Patients Undergoing Total Hip or Knee Arthroplasty. Available at: http://www.aaos.org/research/guidelines/pe_guideline.pdf. Accessed 11 Mar 2011
- 3.U.S. Department of Health and Human Services (2008) The Surgeon General’s Call to Action to Prevent Deep Vein Thrombosis and Pulmonary Embolism. Available at: http://www.surgeongeneral.gov/topics/deepvein/calltoaction/call-to-action-on-dvt-2008.pdf. Accessed 11 Mar 2011
- 4.Editorial Committee on Japanese Guideline for Prevention of Venous Thromboembolism (2004) Japanese Guideline for Prevention of Venous Thromboembolism. Medical Front International, Tokyo (in Japanese)Google Scholar
- 5.The Japanese Circulation Society (2009) Guidelines for the Diagnosis, Treatment and Prevention of Pulmonary Thromboembolism and Deep Vein Thrombosis. Available at: http://www.j-circ.or.jp/guideline/pdf/JCS2009_andoh_h.pdf. Accessed 20 Feb 2011 (in Japanese)
- 6.The Agency for Healthcare Research and Quality (2011) Available at: http://www.qualityindicators.ahrq.gov/. Accessed 1 June 2011
- 7.Eikelboom JW, Karthikeyan G, Fagel N et al (2009) American Association of Orthopedic Surgeons and American College of Chest Physicians guidelines for venous thromboembolism prevention in hip and knee arthroplasty differ: what are the implications for clinicians and patients? Chest 135:513–520PubMedCrossRefGoogle Scholar
- 8.Kuroiwa M, Furuya H, Seo N et al (2010) Incidence and clinical characteristics of perioperative pulmonary thromboembolism in Japan in 2008: results from the annual study of Japanese Society of Anesthesiologists, Committee on Patient Safety and Risk Management, Perioperative Pulmonary Thromboembolism Working Group]. Masui 59:667–673 (in Japanese)PubMedGoogle Scholar
- 11.The Joint Commission (2010) Venous Thromboembolism (VTE) Core Measure Set. Available at: http://www.jointcommission.org/. Accessed 1 Mar 2011