ABSTRAK
Penduduk berusia lanjut (usila) di Indonesia diproyeksikan bertambah jumlahnya di masa mendatang, karena meningkatnya umur harapan hidup. Masalah yang sering dijumpai pada populasi usila tersebut adalah dehidrasi, yaitu tubuh kekurangan cairan dan elektrolit tertentu. Namun keadaan dehidrasi khusus pada usila ini sulit dideteksi, karena tanda dan gejala klinis sering tidak jelas, sehingga dapat berakibal fatal sampai dengan kematian. Dengan beranjak pada prinsip add lifes to years, yang berharap turunnya angka morbiditas pada penduduk usila, termasuk keadaan dehidrasi dan akibatnya, maka dirasakan perlu membuat model matematis yang praktis yang dapat digunakan untuk mendeteksi dini dehidrasi pada pasien usia lanjut berdasarkan variabel pemeriksaan klinis.Model menggunakan acuan baku (gold standard) pemeriksaan laboratorium tertentu, dengan prediktor adalah tanda dan gejala klinis yang relatif lebih mudah dan murah. Penelitian dilakukan di Rumah Sakit C:ipto Mangunkusumo Jakarta, dengan pemeriksaan laboratorium dilakukan oleh Sub-bagian Geriatri Bagian Penyakit Dalam, serta laboratorium Prodia dan Rumah Sakit Harapan Kita Jakarta. Dengan berbagai kendala, akhirnya sebagian pasien yang memenuhi kriteria lelah dikumpulkan datanya, dan dianalisis. Pcrsamaan regresi linier menunjukkan bahwa beberapa tanda dan gejala klinis patut diperhitungkan dalam mendeteksi dehidrasi pada usia lanjut.

---

ABSTRACT

The number of elderly people in Indonesia is projected to be bigger since life expectancy increases. The most frequent problem faced among them is dehydration, which is a condition that the body is sulicred from the lack of water and certain clectrolite. This condition is difficult to be detected because the unobvious clinical sign and symptoms, so it could entail in fatal stage or death'. Based on the principal terms "add lire to years", we do hope that the morbidity in elderly people and its outcome are able to be lowered. Therefore a kind of tool used to detect dehydration as early as possible is needed, especially in clinics in remote areas. This tool is expected to be easy and cheap to be applied, since it is based on the anamneses and physical examination.

The development of mathematical model is relied on the gold standard on laboratory examination, which is predicted by clinical sign and symptom. The study was conducted in Cipto Mangunkusumo Hospital Jakarta (RSCM). The laboratory exam was done by Sub Department of Geriatry, Internal Department, RSCM, and helped also come from Prodia and Harapan Kita Hospital Jakarta. Constraint during data collection were happened, but part of the patient's data finally were able to be gathered and analyzed. Linear regression analysis revealed that some clinical sign and symptom should be considered in diagnosing dehydration among elderly people.