ABSTRAK
Latar belakang: Stratifikasi risiko terhadap pembedahan sangat membantu dalam pengambilan keputusan klinis perioperatif, edukasi, evaluasi, dan audit klinis. Kraniotomi pada tumor otak sebagai tindakan pembedahan berisiko tinggi belum memiliki stratifikasi risiko yang akurat di RSUPNCM karena masih menggunakan ASA yang bersifat subjektif dan kurang informatif. P-POSSUM terbukti tepat dalam prediksi mortalitas pascabedah kraniotomi di India dan Inggris, namun belum diketahui ketepatannya di Indonesia, khususnya di RSUPNCM. Tujuan: Penelitian ini bertujuan untuk mengetahui ketepatan P-POSSUM dalam prediksi mortalitas pascabedah kraniotomi pada tumor otak di RSUPNCM. Metode: Disain penelitian adalah deskriptif analitik retrospektif terhadap seluruh pasien dewasa dengan tumor otak yang menjalani kraniotomi di RSUPNCM selama periode Januari 2015 - Desember 2016. Hasil: Sebanyak 196 subjek dilibatkan dalam analisis risiko mortalitas. Didapatkan rasio O:E 1,68 secara keseluruhan dengan rasio O:E 1,91 pada jangkauan risiko 0-5 dan 1,69 pada jangkauan risiko 11-20 . Hasil uji Hosmer-Lemeshow menunjukkan perbedaan yang signifikan antara angka mortalitas prediksi dan aktual p=0,006 . Simpulan: P-POSSUM tidak tepat dalam prediksi mortalitas pascabedah kraniotomi di RSUPNCM. Diperlukan kajian dan penyesuaian lebih lanjut sebelum P-POSSUM dapat digunakan pada populasi bedah saraf di RSUPNCM.

---

ABSTRACT
Background Risk stratification in surgery helps in perioperative clinical decision making, education, evaluation, and clinical audit. Craniotomy on brain tumor as a high risk surgery does not have an accurate risk stratification in RSUPNCM because they still use ASA, which is subjective and not informative. P POSSUM had been proven to be accurate in predicting postoperative mortality after craniotomy in India and England, but it has not been studied in Indonesia, especially in RSUPNCM. Aim This study was done to gain knowledge about the accuracy of P POSSUM for predicting mortality after craniotomy in brain tumor in RSUPNCM. Methods This was a retrospective descriptive analytic study on adults with brain tumor scheduled to have elective craniotomy in RSUPNCM between January 2015 ndash December 2016. Result 196 subjects were analyzed in this study. Overall O E ratio was 1.68 with O E ratio of 1.91 in the risk range of 0 5 and 1.69 in the risk range of 11 20 . Hosmer Lemeshow test showed significant difference between predicted and actual mortality rate p 0.006 . Conclusion P POSSUM was not accurate for predicting mortality after craniotomy in RSUPNCM. Further studies and adjustments are needed before P POSSUM can be used in neurosurgery population in RSUPNCM.