Electronic Theses and Dissertation

PEMODELAN PREDIKSI KEJADIAN STUNTING
PADA ANAK BALITA DI KABUPATEN PIDIE

Putri Ilham Sari

ABSTRAK
Latar Belakang. Stunting merupakan permasalahan kesehatan anak yang sangat kompleks. Stunting tidak hanya disebabkan oleh status gizi yang kurang, namun juga dapat disebabkan oleh konteks lainnya muali dari tingkat individu, rumah tangga dan masyarakat. Anak yang stunting merupakan hasil dari masalah gizi kronis sebagai akibat dari makanan yang tidak berkualitas, ditambah dengan morbiditas, penyakit infeksi, dan masalah lingkungan. Sehubungan dengan adanya penetapan lokus fokus intervensi mengenai pencegahan stunting berdasarkan Keputusan Menteri Perencanaan Pembangunan Nasional tahun 2021 tentang Penetapan Perluasan Kabupaten/Kota Lokasi Fokus Intervensi, di Kabupaten Pidie terdapat 13 lokus kejadian stunting. Berdasarkan hasil survey data stunting yang telah peneliti dapatkan, maka peneliti memilih lima kecamatan yang menjadi lokasi khusus stunting di Kabupaten Pidie. Penelitian ini bertujuan untuk melihat determinan dari tiga faktor terhadap kejadian stunting dan menemukan pemodelan prediksi kejadian stunting di lima lokasi khusus Kabupaten Pidie. Metodelogi: Metode yang digunakan dalam penelitian ini adalah Structural Equation Modeling Partial Least Square dan Geographic Information Systems (SEM PLS-GIS) Instrumennya menggunakan kuesioner dengan teknik total sampling dari setiap kecamatan. adapun sampel dari kecamatan Tangse sebanyak 211, Tiro sebanyak 46, Kembang Tanjong sebanyak 321, Simpang Tiga 361 responden dan Mutiara Timur 49 responden. Hasil analisis SEM PLS ditemukan bahwa kejadian stunting di Tangse sesuai dengan perpaduan dua model yaitu Consep Chilhood Stunting dan CSDH yakni faktor individu p value 0,000 dan masyarakat p value 0,006 semuanya memberikan pengaruh yang signifikan terhadap kejadian stunting dengan nila R2 (79,1%) artinya pengaruh dari variabel etrsebuti sangat kuat. Di Tiro hasilnya sama dengan perpaduan dua konsep WHO dimana faktor individu p value 0,000 dan masyarakat p value 0,002 semuanya memberikan pengaruh yang signifikan terhadap kejadian stunting dengan nila R2 (59,7%) pengaruhnya sedang. Sementara untuk tiga kecamatan lain yang di dataran rendah yaitu Mutiara Timur hanya faktor individu p value 0,000 yang signifikan terhadap kejadian stunting dengan nila R2 (18,6%), pengaruhnya lemah. Kecamatan Simpang Tiga hanya faktor individu p value 0,000, rumah tangga p value 0,000 keduanya signifikan terhadap kejadian stunting dengan nila R2 (83,5%) pengaruhnya sangat kuat. Kecamatan Kembang Tanjong faktor individu p value 0,000, rumah tangga p value 0,000 keduanya signifikan terhadap kejadian stunting dengan nila R2 (83,8%)pengaruhnya sangat kuat. Berdasarkan hasil Multi-Group Analysis (MGA) SEM PLS bahwa dari 5 kecamatan yang digunakan dalam menganalisis kasus stunting, maka diperoleh hasil prediksi pemodelan kejadian stunting dipengaruhi oleh variabel individu P value 0,001 dan masyarakat P value 0,000 secara signifikan di Kabupaten Pidie. Berdasarkan analisis MGA, model prediksi kejadian stunting di Kabupaten Pidie selaras dengan konsep Conceptual Childhood Stunting dan Conceptual Social Determinants of Health (CSDH). Besarnya pengaruh Nilai R Square (atau R²) menunjukkan di Kabupaten Pidie model prediksi yang mempengaruhi kejadian stunting yakni sebesar 0,732 sebesar 73,2% Berdasarkan analisis SEM GIS bahwa sebaran kasus tertinggi ada di kecamatan Kembang Tanjong sebanyak 184 kasus dan terendah di Kcamatan Tiro sebanyak 25 kasus. Kesimpulan. Prevalensi kasus stunting di Kabupaten Pidie terutama dipengaruhi oleh karakteristik individu dan masyarakat. Pemerintah Kabupaten Pidie harus melaksanakan intervensi pencegahan stunting yang disesuaikan dengan penyebab spesifik stunting di maisng-masing daerah untuk memastikan semua program intervensi secara efektif menyasar populasi yang tepat.
Kata Kunci: Pemodelan Stunting, Individu, Rumah Tangga, Masyarakat, Pidie

PREDICTION MODELING OF STUNTING INCIDENTS IN TODDLER CHILDREN OF PIDIE REGENCY Putri Ilham Sari ABSTRACT Background. Stunting is a very complex child health problem. It is influenced not just by inadequate nutritional conditions but also by various other factors, from the individual level to household and community contexts. Stunted children result from chronic nutritional issues caused by poor-quality food, along with morbidity, infectious diseases, and environmental factors. Based on the 2021 Decree of the Minister of National Development Planning on the Determination of Regency/City Expansion of Stunting Intervention Focus Locations, there are 13 loci of stunting incidences in Pidie Regency. Researchers identified five sub-districts in Pidie Regency that are critical areas for stunting. This study aims to examine the determinants of three factors related to stunting incidence and develop a predictive model for stunting occurrence in these specific locations. Methodology: The method used in this study was Structural Equation Modeling Partial Least Square and Geographic Information Systems (SEM PLS-GIS). The instrument used a questionnaire with a total sampling technique that included 988 mother and child respondents. The sample from the Tangse sub-district was 211, Tiro was 46, Kembang Tanjong was 321, Simpang Tiga had 361 respondents, and Mutiara Timur had 49 respondents. Result: The results of the SEM PLS analysis indicated that the incidence of stunting in Tangse aligns with two models: the Childhood Stunting Concept and the Commission on Social Determinants of Health (CSDH). The analysis revealed that individual factors had a p-value of 0.000 and community factors had a p-value of 0.006, both showing a significant effect on the incidence of stunting. The R² value was 79.1%, indicating a strong explanatory power of the models. In Tiro, the results are a combination of two WHO concepts, where individual factors p-value of 0.000 and community p-value of 0.002 all significantly influence the incidence of stunting with a value of R2 (59.7%); the influence is moderate. In the lowland sub-districts, particularly Mutiara Timur, only the individual factor shows a statistically significant p-value of 0.000 regarding the incidence of stunting, with an R2 value of 18.6%, indicating a weak effect. Regarding the Simpang Tiga sub-district, only individual factors with a p-value of 0.000 and household p-value of 0.000 are significant on the incidence of stunting with an R2 value (83.5%), and the effect is very strong. Kembang Tanjong sub-district has an individual factor p-value of 0.000 and a household p-value of 0.000, both significant to the incidence of stunting with an R2 value (83.8%) and a very strong influence. Based on the results of the Multi-Group Analysis (MGA) SEM PLS, from the 5 sub-districts used in analyzing stunting cases, the results obtained showed that the prediction of stunting incident modeling was significantly influenced by individual variables P value 0.001 and community P value 0.000 in Pidie Regency. Based on the MGA analysis, the prediction model for the incidence of stunting in Pidie Regency aligns with the concepts of Conceptual Childhood Stunting and the Conceptual Social Determinants of Health (CSDH). The R Square (R²) value indicates that the prediction model accounts for 73.2% of the factors influencing the incidence of stunting, with an R² value of 0.732. Based on the SEM GIS analysis, the highest case distribution is in the Kembang Tanjong sub-district, with 184 cases, and the lowest is in the Tiro sub-district, with 25 cases. Conclusion: The prevalence of stunting cases in Pidie Regency is primarily influenced by the characteristics of individuals and the community. The Pidie Regency government should implement stunting prevention interventions that are tailored to the specific causes of stunting in each region to ensure that all intervention programs effectively target the right population. .Keywords: Stunting Modeling, Individual, Household, Community, Pidie