Evaluation of a clinical prediction model by Wells et al. in the diagnosis of deep venous thrombosis of the lower limbs

Abstracts

CONTEXTO: A aplicação de uma estratégia baseada em um modelo clínico associado ao mapeamento dúplex (MD) pode permitir um diagnóstico da trombose venosa profunda (TVP) mais seguro, eficaz e custo-efetivo. OBJETIVO: Testar o modelo clínico de Wells et al. associado ao MD e verificar a ocorrência de TVP nos pacientes categorizados quanto à probabilidade de apresentar a doença, e determinar se, a partir dos resultados obtidos, seria possível reduzir o número de exames seriados com o MD. MÉTODOS: Os pacientes com suspeita clínica de TVP foram categorizados quanto à apresentação de TVP em baixa, moderada e alta probabilidade (BP, MP, AP) e, em seguida, submetidos ao MD. Pacientes com MD negativo repetiram o exame em 24-48 horas e em 7 dias. Pacientes com exame positivo para TVP foram tratados. Todos os pacientes sem TVP foram convocados para reavaliação clínica em 3 meses. RESULTADOS: A ocorrência de TVP entre os 489 pacientes avaliados foi de 39,1% (191), sendo 35,6% identificados no exame inicial e 3,5% no exame seriado. Os índices de pacientes que apresentaram TVP foram de 6,1% no grupo de BP, 26,9% no grupo de MP e 79,5% no grupo de AP. No exame seriado, o percentual de TVP foi de 2,4, 7,8 e 15,1% nos grupos BP, MP e AP, respectivamente. Dos pacientes com MD negativo, 62,4% compareceram após 3 meses, e piora dos sintomas foi apresentada por apenas um paciente. Neste, o MD mostrou TVP de veia poplítea. CONCLUSÃO: Os resultados obtidos sugerem que, para os pacientes com BP para TVP e MD negativo, seria possível prescindir do exame seriado, devido à baixa ocorrência de TVP neste grupo, tornando, assim, a abordagem diagnóstica mais simples.

Trombose venosa; diagnóstico; ultra-som

ORIGINAL ARTICLE

Evaluation of a clinical prediction model by Wells et al. in the diagnosis of deep venous thrombosis of the lower limbs

Veronica Barreto Fortes^I; Hamilton Almeida Rollo^II; Archangelo Tarciso Fortes Jr.^I; Marcone de Lima Sobreira^III; Fernanda Cardoso Santos^IV; Mariângela Giannini^II; Francisco Humberto de Abreu Maffei^V

^I

^IIPhysician. Assistant professor, Vascular Surgery, FMB, UNESP, Botucatu, SP, Brazil

^IIIPhysician, Vascular Surgery Service, Hospital das Clínicas da FMB, UNESP, Botucatu, SP, Brazil

^IVBiomedical scientist. Vascular Laboratory, Hospital das Clínicas da FMB, UNESP, Botucatu, SP, Brazil

^VProfessor of Vascular Surgery, FMB, UNESP, Botucatu, SP, Brazil

^{Correspondence}

ABSTRACT

BACKGROUND: The application of a diagnostic strategy based on a clinical model associated with duplex scanning (DS) may allow for a safer and more effective/cost–effective diagnosis of deep venous thrombosis (DVT).

OBJECTIVE: To evaluate the clinical model proposed by Wells et al. associated with DS and verify the occurrence of DVT in patients divided into probability of presenting the disease, and assess the possibility of reducing the number of repeated DS based on the results obtained.

METHODS: Suspected DVT patients were accordingly categorized into groups of low, moderate and high DVT probability (LP, MP and HP). The patients were then submitted to DS and those without DVT were rescheduled to repeat the examination in 24–48 hours and in 7 days. Patients positively diagnosed with DVT received proper treatment. All patients without DVT were summoned to return within 3 months.

RESULTS: The incidence of DVT among all 489 patients was 39.1% (191); of these, 35.6% were identified in the first examination and 3.5% in the follow–up. Among patients categorized as LP the occurrence was 6.1%, 26.9% in the MP group and 79.5% in the HP group. On the follow–up exams the incidence of DVT in LP, MP and HP groups was 2.4, 7.8 e 13.2%, respectively. Among patients with negative DS, 62.4% attended the reevaluation in 3 months and only one presented worsening of symptoms. This patient was then diagnosed with popliteal DVT using DS.

CONCLUSION: The results suggest that for patients with LP for DVT and negative DS, follow–up exams are not needed, since the occurrence of DVT was low in this group. This procedure simplifies the diagnostic process.

Keywords: Venous thrombosis, diagnosis, ultrasonics.

Introduction

Deep venous thrombosis (DVT) is a frequent and potentially fatal disease. It is estimated that, in the USA, DVT occurs in 1% of the population per year, and it is the third most common cardiovascular disease.¹

Its most feared complication due to morbidity and mortality is pulmonary embolism (PE), which can be avoided with early diagnosis and anticoagulant treatment. Postthrombotic syndrome is another complication, nonfatal but with a strong socioeconomic impact. It can be present despite a proper DVT treatment. However, when the latter is performed earlier, it tends to minimize the effects of postthrombotic syndrome.

The anticoagulant treatment, despite being highly efficacious in preventing the progression of thrombosis, embolization and recurrent disease, is associated with complications and implies blood tests and regular visits to the physician. It should only be started after diagnostic confirmation of DVT.

The clinical diagnosis of DVT is not accurate, since, on the one hand, only 20–40% of patients with suggestive symptoms have the disease confirmed by objective examinations. On the other hand, 15–50% of DVT cases do not present a characteristic initial clinical status. Therefore, its diagnosis is certainly based on complementary examinations.

Among available diagnostic methods, duplex scanning (DS) is considered the gold standard of noninvasive examinations, presenting good sensitivity and specificity for proximal DVT. However, such accuracy is reduced for distal DVT (leg veins), with sensitivity and specificity around 70% in that segment.² The risk of pulmonary embolism caused by isolated DVT of leg veins seems small, but there is a risk of progression of distal thrombosis for proximal segments up to 20%.³ For that reason, some authors believe that, in cases of negative DS for DVT, it should be repeated every 7 days with the aim of detecting thrombi in progression.^4–7

In an attempt to simplify the diagnostic approach, Wells et al.⁸ developed a model of clinical prediction that classifies patients as to their risk of presenting DVT. This model, associated with noninvasive complementary examinations, proved to be useful in several studies in the diagnostic procedure of DVT.^3,8,9

This study aims at testing the model of clinical prediction by Wells et al. associated with DS in patients with clinical suspicion of DVT receiving care at Hospital das Clínicas da Faculdade de Medicina de Botucatu and verifying: 1) the prevalence of DVT in patients classified as having high, moderate and low probability of presenting the disease; and 2) whether, based on the results obtained, it would be possible to reduce the number of serial studies with DS.

Methods

The research project was approved by the Research Ethics Committee at Faculdade de Medicina de Botucatu da Universidade Estadual Paulista (UNESP).

The patients with clinical suspicion of lower limb DVT hospitalized in wards and Emergency Rooms or Screening Units at Hospital das Clínicas da Faculdade de Medicina de Botucatu, UNESP were prospectively assessed by applying the study protocol.

Inclusion criteria were patients 18 years or older with suspicion of first episode of DVT, whose complaints had started up to 30 days ago. These were included after signing a consent form.

Exclusion criteria were not agreeing to participate in the study, presenting complaints of less than 30 days, presenting signs and symptoms of pulmonary embolism, presenting DVT antecedent, pregnancy, aged less than 18 years and/or being under anticoagulant therapy.

After performing anamnesis, questionnaire about several devices, personal and family history and general and vascular physical examination, a protocol based on the proposal by Wells et al.⁸ was completed. The patient was instructed about the study and requested to read and, in case of agreement, sign a consent term. The assessments were performed by Vascular Surgery residents, first and second year, or by the vascular surgeon on duty, who had been previously instructed about how to complete the protocol. Before starting the study, a 6–week training was carried out to complete the protocol, and the authors checked and corrected the forms.

The protocol included questions to patients or escorts about risk factors to the development of venous thromboembolic disease and the data obtained in the vascular physical examination (Table 1). The score proposed by Wells et al. was adopted, in which the patient receives 1 point for the presence of each of the factors listed in Table 1. In addition, the presence of other diseases that could justify the clinical status was investigated. In case the patient presented a more likely differential diagnosis, 2 points were subtracted from their score (Table 1).

click to enlarge

According to the scores obtained in this protocol, the patients were classified as to probability of having DVT into three groups:

–Patients with low probability (LP) of DVT: equal or lower than 0 points.

–Patients with moderate probability (MP) of DVT: 1 to 2 points.

–Patients with high probability (HP) of DVT: 3 or more points.

Next, the patients were referred to the Laboratory of Noninvasive Methods of Vascular Surgery to undergo DS. Those with positive result for DVT were given proper treatment. In case the result was negative or doubtful for DVT, the examination was repeated in 24–48 and after 7 days. In case there was still doubt, the patient was submitted to phlebography. The patients with negative serial examinations were advised to inform the medical team in case they presented signs or symptoms suggesting DVT or PE and to return in 3 months for a clinical reevaluation (Figure 1). The DS was performed at the Vascular Laboratory using a Platinum Philips CVI (Color Velocity Imaging), with a 7.5 MHz frequency transducer. Whenever necessary, in patients more obese and in deeper veins, a 3 MHz transducer was used, since these were the available transducers at the Laboratory. The examinations were carried out by skilled vascular surgeons or by vascular surgery residents under supervision of the former. All examinations were checked by a professional specialized in vascular ultrasonography.

The venous DS was performed with the patient in a horizontal inclined supine position. A mapping of the deep veins was carried out in all their extension in a cross section, observing their morphology, anatomical distribution, compressibility and presence of venous flow. The external iliac, common femoral, superficial femoral, popliteal, fibular and posterior tibial arteries were assessed. To assess the popliteal and fibular veins, the patient was placed in a horizontal or side prone position with slight knee flexion.

Diagnostic criteria were the compressibility test, verification of venous flow and its changes and, whenever possible, visualization of echogenic material, suggesting thrombus, inside the veins. DVT was considered proximal when it involved popliteal and/or more proximal veins (iliac, common femoral and superficial femoral) and distal when it was restricted to the leg veins (fibular and/or posterior tibial veins).

The patients who returned after 3 months were clinically reassessed and asked about improvement in signs and symptoms of DVT and about complaints related to thromboembolic disease. Next, a vascular physical examination was performed and, in case of worsening of signs and symptoms, a new DS was performed.

For the statistical analysis, normal distribution was used to compare proportions involving two populations, and in case of more than two populations, Tukey's method was used. In case of multinomial proportions, Goodman's test was used.⁹ Significance level was set in 5% (p < 0.05).

Results

We assessed 489 consecutive patients who met the inclusion criteria between August 2000 and December 2002. There were 216 male (44.2%) and 273 female (55.8%) patients, aged between 18 and 96 years, mean age of 57.1 years. Of all assessed patients, 291 (59.3%) were outpatients and 198 (41.7%) were inpatients, including clinical and surgical patients, independently of hospitalization time.

There were 147 (30%) patients with LP, 171 (35%) with MP and 171 (35%) with HP.

DVT was present in 191 patients (39.1%), and DVT in groups of LP, MP and HP occurred in 6.1% (9), 26.9% (46) and 79.5% (136), respectively. The difference is statistically significant (Table 2). The occurrence of DVT was higher among men (44.9%) compared with women (34.4%), and it was a statistically significant difference.

Click to enlarge

There was no significant difference between inpatients (40.7%) and outpatients (38.1%) with DVT. Among outpatients, the occurrence of DVT in LP, MP and HP groups was 5.6, 31.2 and 86.2%, respectively. Among inpatients, such occurrence was 7.7% in the LP group, 21.3% for MP and 72.6% for HP. When comparing the occurrence of DVT in inpatients and outpatients, there was statistical difference between patients with HP, with prevalence of DVT in outpatients (Table 3).

Click to enlarge

There was higher occurrence of proximal DVT compared with distal DVT in MP and HP groups. Among patients with LP, there was no statistically significant difference. The percentage of proximal DVT was not different between LP and MP groups (55.6 and 69.6%), but it was higher in the HP group (90.4%) (Table 4).

Click to enlarge

Of the 298 patients with initial negative DS, 273 (91.6%) returned to perform the serial examinations; 124 (87.9%) patients in the LP group, 116 (92.8%) in the MP group and 33 (95%) in the HP group. There was statistically significant difference when the number of visits in the LP group was compared with the number of visits in MP and HP groups. DVT was diagnosed in 8.9% (17) of these patients: 2.4% (3) in the LP group, 7.8% (9) in the MP group and 15.1% (5) in the HP group. There was statistical difference for the occurrence between the three groups (Table 5). Two patients died before the end of the study due to causes that are probably unrelated to DVT (one stroke and one head trauma). However, necropsy was not authorized.

Click to enlarge

Of the 298 patients without DVT detected in serial examinations, 186 (62.4%) had another visit in 3 months. The percentage of thromboembolic complications was 0.5%. Of the 186 patients, 47.8% were in the LP group, and none presented complaints suggesting thromboembolic disease. The percentage of visits in MP and HP groups was statistically different in the LP group. Of the patients with MP, 77.6% had another visit and one patient presented worsening in signs and symptoms. A DS was performed, showing DVT of the popliteal and posterior tibial veins. The other patients did not present complaints. The rate of visits in the HP group was statistically similar to the MP group (65.7%), and no patients presented complaints compatible with DVT or PE (Table 6). In this group, four patients died before the assessment. Two patients due to sepsis, one due to malignant neoplasia and one due to opportunistic infection caused by HIV infection. Necropsy was not performed in any case. Therefore, it was not possible to exclude thromboembolic disease.

Click to enlarge

Discussion

We found a similar distribution of the 489 patients included in the study between LP, MP and HP groups, with 30, 35 and 35%, respectively. This distribution is different from that reported by Wells et al. in their 1995 and 1997 studies, in which patients with LP were prevalent, with 55.5 and 57%, respectively.^10,11 Wells et al. did not include inpatients in their sample, which could explain the difference in distribution. However, even when we considered only the outpatients in our sample, there was no statistical difference, although we had found a slight prevalence of patients with LP (37%), compared with MP (33%) and HP (30%) groups. This could be explained by the fact that, in our country, patients are referred later for assessment, when the signs and symptoms are more important. This fact could reflect a delay in referring patients by general physicians or a tendency patients have of seeking the physician only when the clinical status is advanced.

The total occurrence of DVT in the sample was 39.1%, which was different from several studies.^10–15 It was 38.1% among outpatients and 40.7% among inpatients. A higher prevalence should be expected among inpatients compared with outpatients. However, this was not the case, probably due to the fact that our sample of outpatients did not exclude those who had been recently hospitalized or were coming from nursing homes, considering that this information was not investigated in the initial assessment. Heit et al.,¹⁶ in a retrospective study assessing the incidence of DVT in patients of a community, reported an occurrence of DVT 100 times higher in inpatients, including symptomatic and asymptomatic, when compared with the general population.

A previous study in our service carried out by Rollo et al.¹⁷ found 51.6% of DVT in 658 patients with clinical suspicion of DVT. Such higher incidence can be explained by a more judicious use of complementary examinations, excluding DS in patients with other diseases that could justify the signs and symptoms.

Wells et al.^10,11 found prevalence of 25.5 and 16%, respectively, in outpatients. Anderson et al.,⁸ assessing patients in the emergency room with suspicion of DVT, found prevalence of 13.1%. Wells et al.¹⁴ found prevalence of 270 and 21.1%, respectively, in outpatients.

In a retrospective study assessing 540 patients with suspicion of DVT submitted to DS, Mattos et al.¹⁸ found DVT prevalence of 24%, and thrombi were restricted to the leg veins in 33% of cases. In our patients, among the 191 cases of DVT, 18.8% were limited to the distal segment, 44.4% to the LP group, 30.4% to the MP group and 9.6% to the HP group. There was statistically significant difference between the HP group and the MP and LP groups, which, however, were not different between themselves, showing the correlation of thrombosis extension with intensity of clinical status.

Souza,¹⁹ in a study using phlebography, reported 25% of DVT in the LP group, 50% in the MP group and 85% in the HP group. Among patients with DVT, he found proximal DVT in 12.5% in the LP group, 25% in the MP group and 80.6% in the HP group. In the present study, we found, among patients with DVT, proximal involvement in 91.1% of patients: 55.6% among patients classified as LP, 69.6% as MP and 90.4% as HP (Table 4). Souza¹⁹ may have found higher frequency of DVT in LP and MP groups than what was observed in this study possibly because he used phlebography as a diagnostic method, which is a more sensitive examination than DS. This also explains why he found higher occurrence of distal DVT than in this study.

The occurrence of DVT in LP, MP and HP groups was 6.1, 26.9 and 79.5%, respectively. Among outpatients it was 5.6, 31.2 and 86.2%, and among inpatients it was 7.7, 21.3 and 72.6%. These findings confirm those by Wells et al.,^10,11 who, in their studies with outpatients, found, in 1995, DVT prevalence of 5, 33 and 85% in LP, MP and HP groups; in 1997, they found 3, 17 and 75%, respectively. Anderson et al.⁸ found similar proportions for LP and MP groups, with 3.2 and 14.3%, respectively, and higher prevalence of DVT for the HP group (49%).

It has been well established that patients with suspicion of DVT should be submitted to DS or to another available complementary examination, preferentially noninvasive.^20.21 The DS is available in most services. This examination can be easily performed and does not cause discomfort for the patient. However, doubt remains for patients with negative DS, since 10–20% of patients with symptomatic DVT have isolated DVT of the leg, which can go unnoticed in the DS;²¹ at least 20–30% of these thrombi will evolve proximally, increasing the risk of pulmonary embolism.^3.22–24 For that, the serial examination is proposed when the DS is negative in symptomatic patients to detect these cases.²⁵ In our study, the serial examination was performed in 91.6% of the patients with negative initial examination. The highest number of return visits was in the HP group (95%), probably because it included more symptomatic patients. In the other groups, it was necessary to perform an active search in a large number of cases, especially among LP patients.

The prevalence of DVT in the serial examination was 8.9%: 2.4% in the LP group, 7.8% in the MP group and 15.1% in the HP group. It is important to stress that the three patients in the LP group who presented positive serial examination also presented symptom worsening, suggesting that patients with LP could dismiss this examination, except in cases of worsening in clinical status. Therefore, it would be possible to reduce DS examinations in around 29%.

Wells et al.,¹⁰assessing 593 patients with suspicion of DVT, performed serial examination in 7 days only in patients with MP. Patients with HP and negative DS were submitted to phlebography, as well as patients with LP and positive DS. They observed 1.8% of DVT in the serial DS. The technique used to assess the deep venous system in the studies by Wells et al. is different from that used in our service. Those authors only assess the proximal veins until the trifurcation of the popliteal vein, without assessing leg veins. This could partly explain the cause of having found a higher number of DVT cases in the serial examination in this study. Among patients with HP and negative DS, Wells et al. found DVT in 23.5% of phlebographies. At the 3–month follow–up of patients with DVT, there were only three thromboembolic events (0.6%); one among patients with LP and two among patients with MP.

We found similar results during the follow–up of our patients, but the number of return visits in 3 months was lower than the visits for serial DS. Of the 186 patients (62.4%) who returned, only one, in the MP group, presented worsening of initial complaints with occurrence of new symptoms. He was submitted to a new DS, which showed DVT of the popliteal vein and posterior tibial veins. No patients returned with complaints suggesting PE. The rate of thromboembolic complications was low (0.5%) over the 3–month follow–up; therefore, this type of follow–up has no justification for most patients.

Anderson et al.,⁸ using the same approach of Wells et al., found 1.1% of DVT in the serial examination and in the HP group with negative DS 13.8% presented DVT in the phlebography. One patient in the LP group had worsening of signs and symptoms and was submitted to DS in 7 days, being diagnosed a DVT of the leg. Those authors observed 0.7% of thromboembolic complications in the 3–month follow–up, a result that is similar to that found in the present study.

The prevalence of DVT in LP, MP and HP groups corresponded to the expectation, i.e., it was low in the LP group and high in the HP group.

In conclusion, the results obtained in this study suggest that the clinical prediction model proposed by Wells et al. associated to DS may facilitate the diagnostic strategy of DVT. They also suggest that, in patients classified as LP and negative DS, it is not necessary to perform the serial DS, since in this group the prevalence of DVT during the follow–up was low. Therefore, the examination should be repeated only when there is worsening of signs and symptoms.

References

Manuscript received June 25, 2006, accepted January 30, 2007.

This study was presented at I Congresso Brasileiro de Eco–Doppler Vascular, held on July 21 to 24, 2004 in Vitória, ES, Brazil.

• 1. Hirsh J, Hoak J. Management of deep vein thrombosis and pulmonary embolism. A statement for healthcare professionals. Council on Thrombosis (in consultation with the Council on Cardiovascular Radiology), American Heart Association. Circulation. 1996;93:2212-45.
• 2. Hirsh J, Raschke R, Warkentin TE, Dalen JE, Deykin D, Poller L. Heparin: mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy and safety. Chest. 1995;108(4 Suppl):258S-75S.
• 3. Cogo A, Lensing AW, Prandoni P, Hirsh J. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med. 1993;153:2777-80.
• 4. Cogo A, Lensing AW, Koopman MM, et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study. BMJ. 1998;316:17-20.
• 5. Kearon C, Julian JA, Newman TE, Ginsberg JS. Noninvasive diagnosis of deep vein thrombosis. McMaster Diagnostic Imaging Practice Guidelines Initiative. Ann Intern Med. 1998;128:663-77.
• 6. Wolf B, Nichols DM, Duncan JL. Safety of a single duplex scan to exclude deep venous thrombosis. Br J Surg. 2000;87:1525-8.
• 7. Hirsh J, Lee A. How we diagnose and treat deep vein thrombosis. Blood 2002;99:3102-10.
• 8. Anderson DR, Wells PS, Stiell I, et al. Thrombosis in the emergency department: use of a clinical diagnosis model to safely avoid the need for urgent radiological investigation. Arch Intern Med. 1999;159:477-82.
• 9. Tick LW, Ton E, van Voorthuizen T, et al. Practical diagnostic management of patients with clinically suspected deep vein thrombosis by clinical probability test, compression ultrasonography, and D-dimer test. Am J Med. 2002;113:630-5.
• 10. Wells PS, Anderson DR, Bormanis J, et al. Value of assessment of pretest probability of deep-vein thrombosis in clinical management. Lancet. 1997;350:1795-8.
• 11. Zar JH. More on dichotomous variables. In: Zar JH. Biostatistical analysis. 4th ed. Upper Saddle River: Prentice Hall; 1999. p. 516-65.
• 12. Wells PS, Hirsh J, Anderson DR, et al. Accuracy of clinical assessment of deep-vein thrombosis. Lancet. 1995;345:1326-30.
• 13. Perrier A, Desmarais S, Miron MJ, et al. Non-invasive diagnosis of venous thromboembolism in outpatients. Lancet. 1999;353:190-5.
• 14. Miron MJ, Perrier A, Bounameaux H. Clinical assessment of suspected deep vein thrombosis: comparison between a score and empirical assessment. J Intern Med. 2000;247:249-54.
• 15. Dryjski M, O'Brien-Irr MS, Harris LM, Hassett J, Janicke D. Evaluation of a screening protocol to exclude the diagnosis of deep vein thrombosis among emergency department patients. J Vasc Surg. 2001;34:1010-5.
• 16. Heit JA, Melton LJ 3rd, Lohse CM, et al. Incidence of venous thromboembolism in hospitalized patients vs community residents. Mayo Clin Proc. 2001;76:1102-10.
• 17. Rollo HA, Lastoria S, Yoshida WB, et al. Mapeamento dúplex no diagnóstico da trombose venosa profunda dos membros inferiores. In: Programa Oficial do 32 Congresso Brasileiro de Angiologia e Cirurgia Vascular; 1997, Curitiba. 120.
• 18. Mattos MA, Melendres G, Sumner DS, et al. Prevalence and distribution of calf vein thrombosis in patients with symptomatic deep venous thrombosis: a color-flow duplex study. J Vasc Surg. 1996;24:738-44.
• 19. Souza E. Validação de modelo de predição clínica para o diagnóstico de trombose venosa profunda dos membros inferiores [dissertação]. São Paulo: Universidade Federal de São Paulo; 2001.
• 20. Haeger K. Problems of acute deep venous thrombosis. I. The interpretation of signs and symptoms. Angiology. 1969;20:219-23.
• 21. Richards KL, Armstrong JD Jr., Tikoff G, Hershgold EJ, Booth JL, Rampton JB. Noninvasive diagnosis of deep-vein thrombosis. Arch Intern Med. 1976;136:1091-6.
• 22. Wells PS, Hirsh J, Anderson DR, et al. A simple clinical model for the diagnosis of deep vein thrombosis combined with impedance plethysmography: potential for an improvement in the diagnostic process. J Intern Med. 1998:243:15-23.
• 23. Kakkar VV, Howe CT, Flanc C, Clarke MB. Natural history of postoperative deep-vein thrombosis. Lancet. 1969;2:230-2.
• 24. Lagerstedt CI, Olsson CG, Fagher BO, Oqvist BW, Albrechtsson U. Need for long-term anticoagulant treatment in symptomatic calf-vein thrombosis. Lancet. 1985;2:515-8.
• 25. Cogo A, Lensing AW, Prandoni P, Hirsh J. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med. 1993;153:2777-80.

Publication Dates

• Publication in this collection
  19 July 2007
• Date of issue
  Mar 2007

History

• Received
  25 June 2006
• Accepted
  30 Jan 2007