<![CDATA[AbstrakSalah satu upaya yang cukup efektif dalam surveilans malaria adalah melakukan screening (penapisan) malaria untuk meningkatkan sistem kewaspadaan dini di kelompok masyarakat daerah endemis malaria. Hasil penapisan positif atau meragukan harus dirujuk ke dokter untuk penegakkan diagnosis dan pengobatan. Penelitian ini bertujuan untuk menganalisis validitas gejala klinis sebagai indicator untuk memprediksi kasus malaria di Indonesia dengan menggunakan disain cross-sectional. Sampel dalam penelitian ini adalah semua responden yang diwawancarai, dilakukan pemeriksaan darah dengan Rapid Diagnostic Test (RDT). Validitas gejala klinis diukur dengan melakukan summary statistic untuk diagnostic test. Di wilayah endemis tinggi sensitivitas gejala klinis demam saja sebagai prediktor malaria hanya 26,9% (95% CI: 22-32,2) dan PPV 11,4% (95% CI: 9,2-13,9) dengan spesifisitas 96% (95% CI: 95,6-96,3). Sensitivitas, PPV, dan spesifisitas gejala demam saja di daerah endemis sedang secara berturut-turut adalah sebesar 26,1% (95% CI: 17,5-36,3), 5.0% (95% CI: 3,2-7,4), dan 96,9% (95% CI: 96,6-97,2). Di daerah endemis rendah sensitivitas demam sebagai alat diagnosa kasus malaria hanya sebesar 3,5% (95% CIH: 1,6-6,6) dengan PPV 1,1% (95% CI: 0,5-2,1) sedangkan spesifisitas 98% (95% CI: 97,8-98,1). Kombinasi gejala klinis demam, menggigil, sakit kepala, berkeringat, mual, dan muntah dalam analisis data Riskesdas menunjukkan sensitifitas yang lebih tinggi bila dibandingkan dengan demam saja (36,4%). Sebaliknya PPV kombinasi gejala tersebut lebih rendah bila dibandingkan dengan PPV demam saja (3,8%). Gejala klinis malaria kurang valid untuk digunakan untuk mendeteksi kasus malaria baik pada daerah endemis tinggi, sedang, maupun rendah. Akan tetapi penggunaannya untuk daerah endemis tinggi masih dimungkinkan, seperti yang direkomendasikan oleh WHO terutama untuk anak-anak.Kata kunci: malaria, gejala klinis, validitas, sensitivitas, spesifisitasAbstractOne of the effective ways in malaria surveillance is screening to improve early warning system in communities in malaria endemic area. Positive screening or doubted results should be referred to physician for diagnosis and treatment. The aim of this study was to analyse the validity of clinical symptoms as an indicator to predict malaria case in Indonesia. Samples of this study were all respondents interviewed by the National Health Research in 2010 whose blood were examined for malaria using RDT. Validity of clinical symptoms was analysed by using summary statistic for diagnostic test. The results showed that the sensitivity of fever alone as a predictor of malaria in high endemic area was only 26.9% (95% CI: 22-32.2). However, the specificity was 96% (95% CI: 95.6-96.3) and positive predictive value (PPV) 11.4% (95% CI: 9.2-13.9). In low endemic area, sensitivity and PPV of fever alone were low with 3.5% (95% CI: 1.6-6.6) and 1.1% (95% CI:0.5-2.1) respectively. On the other hand, the specificity was relatively high with 98% (95% CI: 97.8-98.1). Combination of fever, chill, headache, sweat, nausea, and vomit showed higher sensitivity (36.4%; 95% CI:28.9-40.5) and specificity (84.2%; 95% CI: 83.6-84.8) compare to fever alone in high endemic area, whereas the PPV was lower (3.8%: 95% CI: 3.1-4.6). In low endemic area, symptoms combination had a higher sensitivity (14.7%; 95% CI: 10.6-19.7) compare to fever alone. However, the specificity and PPV were lower with 91.2% (95% CI: 90.9-91.4) and 1% (95% CI: 0.7-1.4) respectively. Conclusion: The validity of clinical signs and symptoms to diagnose malaria is low in high and low transmission area. However, the use of clinical symptoms as a predictor of malaria is still possible in high transmission area as recommended by WHO, particularly for children.Key words: malaria, clinical symptoms, validity, sensitivity, specificity]]>
