UJI AKURASI MODEL KLINIS PRETES PROBABILITAS DUGAAN TROMBOSIS VENA DALAM (TVD) DARI WELLS DI RSUP Dr SARDJITO

Dwi Wulansih, dr. Johan Kurnianda, SpPD-KHOM

2012 | Tesis | S2 Ilmu Penyakit Dalam

Latar Belakang : Diagnosis trombosis vena dalam/TVD merupakan masalah penting di bidang kedokteran karena sejumlah besar kasus tidak terdiagnosis. Diagnosis TVD secara akurat sangat penting untuk mencegah komplikasi akut emboli paru dan komplikasi jangka panjang berupa sindroma postphlebitis dan hipertensi pulmonar serta untuk menghindari terapi antikoagulan yang tidak tepat yang dihubungkan dengan risiko tinggi perdarahan. Terdapat beberapa model klinis dugaan TVD antara lain skor Hamilton dan skor Wells. Model klinis pretes probabilitas dugaan TVD dari Wells telah banyak digunakan dan teruji validitasnya di beberapa negara di Eropa. Uji akurasi model klinis pretes probabilitas dugaan TVD dari Wells belum pernah dikerjakan untuk ras Asia dan di Indonesia. Tujuan Penelitian : Penelitian ini bertujuan untuk menguji akurasi, presisi (sensitivitas, spesifisitas dan nilai duga) model klinis pretes probabilitas dugaan TVD dari Wells untuk mendiagnosis TVD pada pasien-pasien di RSUP Dr Sardjito. Metode : Desain penelitian diagnostik ini menggunakan desain potong lintang, mengikutkan pasien-pasien yang mengalami gejala TVD di poliklinik Tulip, rawat jalan dan rawat inap di RSUP Dr Sardjito selama bulan Januari 2011-Mei 2012. Sebanyak 100 pasien yang ikut dalam penelitian ini dilakukan pemeriksaan model klinis pretes probabilitas dugaan TVD dari Wells dan diperiksa USG doppler dan D-Dimer. Hasil USG doppler yang pertama negatif akan dilakukan USG ulang 1 minggu kemudian. Hasil : Sebanyak 56 pasien (56 %) menderita TVD dari 100 subyek penelitian. Karakteristik dasar penelitian menunjukkan jumlah wanita lebih banyak dibanding pria dan rerata usia 52,98 ± 1,335 tahun. Faktor risiko TVD dengan frekuensi paling banyak adalah kanker 49 pasien (49%), bed ridden 42 pasien (42 %), usia lanjut 33 pasien (33 %). Akurasi model klinis pretes probabilitas dugaan TVD dari Wells sebesar 81 %, presisi 77,61 %, sensitivitas tinggi 92,86 %, spesifisitas kurang yaitu 65,91 %, nilai duga negatif 87,88 %, rasio kemungkinan negatif 0,11 dan prevalensi 56 %. Simpulan : Model klinis pretes probabilitas dugaan TVD dari Wells mempunyai akurasi tinggi, presisi sedang, sensitivitas tinggi dan spesifisitas kurang untuk mendiagnosis pasien-pasien di RSUP Dr Sardjito.

Background: The diagnosis of deep vein thrombosis / DVT is an important issue in the medicine because a large number of cases are undiagnosed. The accurate diagnosis of the DVT an is essential to prevent complications of acute pulmonary embolism and long-term complications such as pulmonary hypertension and postphlebitis syndrome and to avoid improper anticoagulant therapy associated with a high risk of bleeding. There are several models of clinical suspicion of DVT include Hamilton’s score and Wells’s score. The Wells clinical model for predicting pretest probability for DVT has been widely used and tested its validity in several countries in Europe. Test the accuracy of a Wells clinical model for pretest probability for DVT has not been done for the races of Asia and in Indonesia. Objective: This study aimed to test the accuracy, precision (sensitivity, specificity and predictive value) of a Wells clinical model for predicting pretest probability for DVT to diagnose of DVT for patients in the Dr. Sardjito Hospital. Methods: The study used a cross-sectional, diagnostic design, include patients who have symptoms of DVT were visited the clinic Tulip, outpatient and inpatient in the Dr Sardjito Hospital during January 2011-May 2012. In this study, a total of 100 patients participated underwent Wells clinical model of pretest probability for DVT examination and underwent doppler ultrasound and D-Dimer examination. The results of negative in the first ultrasound examination will be examined again 1 week later. Results: DVT was confirmed in 56 patients (56%) from 100 patients. The baseline characteristics showed the number of women more than men and the mean age of 52.98 ± 1.335 years. The most frequency for risk factors of DVT were cancer 49 patients (49%), bed ridden 42 patients (42%), older 33 patients (33%). Accuracy of the Wells clinical model of pretest probability for DVT was 81%, precision was 77,61 %, high sensitivity was 92.86%, specificity was 65.91%, negative predictive value was 87.88%, negative likelihood ratio was 0.11 and prevalence was 56%. Conclusion: The Wells clinical model of pretest probability for DVT has a high accuracy, moderate precision, high sensitivity and poor specificity to diagnosed of DVT in the Dr. Sardjito Hospital.

Kata Kunci : trombosis vena dalam/TVD, model klinis pretes probabilitas dugaan TVD dari Wells, uji akurasi, presisi