Abstrak: Outcome misclassification is widespread in epidemiology, but methods to account for it are rarely used. We describe the use of multiple imputation to reduce bias when validation data are available for a subgroup of study participants. This approach is illustrated using data from 308 participants in the multicenter Herpetic Eye Disease Study between 1992 and 1998 (48% female; 85% white; median age, 49 years). The odds ratio comparing the acyclovir group with the placebo group on the gold-standard outcome (physician-diagnosed herpes simplex virus recurrence) was 0.62 (95% confidence interval (CI): 0.35, 1.09). We masked ourselves to physician diagnosis except for a 30% validation subgroup used to compare methods. Multiple imputation (odds ratio (OR) = 0.60; 95% CI: 0.24, 1.51) was compared with naive analysis using self-reported outcomes (OR = 0.90; 95% CI: 0.47, 1.73), analysis restricted to the validation subgroup (OR = 0.57; 95% CI: 0.20, 1.59), and direct maximum likelihood (OR = 0.62; 95% CI: 0.26, 1.53). In simulations, multiple imputation and direct maximum likelihood had greater statistical power than did analysis restricted to the validation subgroup, yet all 3 provided unbiased estimates of the odds ratio. The multiple-imputation approach was extended to estimate risk ratios using log-binomial regression. Multiple imputation has advantages regarding flexibility and ease of implementation for epidemiologists familiar with missing data methods.

Abstrak: Pendahuluan. Indonesia merupakan negara yang masih banyak layanankesehatannya terletak di daerah perifer dengan fasilitas minim dan jarang memilikitenaga ahli untuk memprediksi berat bayi saat dilahirkan.Metode. Penelitian ini menggunakan desain studi cross sectional. Kriteria inklusiibu melahirkan anak terakhir, bayi lahir hidup, dan bayi tunggal, didapatkan sampelsebanyak 23.689.Hasil. Variabel yang menjadi faktor risiko kejadian BBLR adalah usia kehamilan(POR 2,01), umur (POR 1,28), paritas (POR 1,56), tinggi ibu (POR 1,48),komplikasi (POR 1,46). Analisis ROC didapatkan area under curve untukmengidentifikasi kejadian BBLR sebesar 0,602. Nilai titik potong untuk skoringprediksi 4 dan sensitivitas 59,8%.Kesimpulan. Usia kehamilan, umur, paritas, tinggi ibu, dan komplikasi merupakanfaktor risiko dan dapat digunakan untuk memprediksi bayi yang akan dilahirkanberisiko BBLR.Kata Kunci: berat lahir bayi, sensitivitas, prediksi

---

Introduction. Indonesia is a country that still many health services located inperipheral areas with minimal facilities and rarely have experts to predict the weightof the baby at birth.Methods. This study using cross sectional study design. The inclusion criteriamaternal last child, a baby was born alive, and a single baby, obtained a sample of23.689.Results. Variables are a risk factor for LBW is gestational age (POR 2,01), age(POR 1,28), parity (POR 1,56), maternal height (POR 1,48) and complications(POR 1,46). ROC analysis obtained an area under the curve to identify the LBW of0,602. Value cut-off point for scoring 4 prediction and sensitivity of 59,8%.Conclusion. Gestational age, age, parity, height, and complications are risk factorsand can be used to predict the baby to be born at risk of LBW.Keywords: birth weight babies, sensitivity, predictive.