Comparison of Different Underlying Distributions in The Accelerated Failure Time (AFT) Model on Mortality of Covid-19 Patients

Asti Meiza, Asep Solih Awaluddin, Nurapni Oktapia Hidayah, Adnan Taha


In 2022, the COVID-19 virus is still making headlines in various mass media because it is a virus that is very dangerous to health. The world health organization, WHO, explained that the virus caused a global pandemic that infected the whole world. The condition of a pandemic has not yet turned into an endemic. Based on the total confirmed COVID-19 positive cases, Indonesia ranks 18th in the world out of 222 infected countries. To determine the influence factors on COVID-19 cases, survival analysis is one of the techniques that could be applied. One of the most commonly used models in survival analysis is Accelerated Failure Time (AFT) model. In the AFT model, it is required to check assumptions regarding the feasibility of the distribution form. In this study, the distributions used are Weibull, Exponential, Log-normal, and Log-logistics distributions. We compare each distribution to get the best model to analyze death cases due to COVID-19. Comparisons are made by comparing the AIC values of each distribution. The best model is selected based on the smallest AIC value. The AFT model with a log-normal distribution is selected as the best model with an AIC value of 142.763. The AIC value for this log-normal distribution is the smallest compared to the AIC value for other distributions.

Keywords: accelerated failure time model; COVID-19; mortality analysis; survival analysis.



Tahun 2022, virus COVID-19 masih menjadi berita utama di berbagai media massa karena merupakan salah satu virus yang sangat berbahaya bagi kesehatan. Badan organisasi kesehatan dunia, WHO menjelaskan bahwa virus menyebabkan  terjadinya pandemi global yang menginfeksi seluruh dunia. Kondisi pandemi masih belum berubah menjadi endemi. Dari total yang terkonfirmasi positif COVID-19 Indonesia menduduki posisi ke-18 di dunia dari 222 negara yang terinfeksi. Untuk mengetahui faktor-faktor yang berpengaruh terhadap COVID-19 dan untuk menentukan model dari COVID-19 ini salah satunya dapat dilakukan dengan analisis survival. Salah satu model survival yang digunakan yaitu model Accelerated Failure Time (AFT).  Dalam model AFT ini diharuskan melakukan pengecekan asumsi-asumsi mengenai kelayakan bentuk distribusi. Pada penelitian ini distribusi yang digunakan yaitu distribusi Weibull, Eksponensial, Log-normal, dan Log-logistik. Dilakukan perbandingan antar tiap distribusi untuk mendapatkan model terbaik yang dapat digunakan dalam menganalisis kasus kematian akibat COVID-19. Perbandingan dilakukan dengan membandingkan nilai AIC dari  setiap distribusi. Hasil penelitian memilih model AFT dengan distribusi log-normal sebagai model terbaik dengan nilai AIC sebesar 142,763. Nilai AIC untuk distribusi log-normal ini paling kecil dibandingkan dengan nilai AIC untuk distribusi lainnya.

Kata Kunci: analisis mortalitas; analisis survival; COVID-19; model accelerated failure time.


2020MSC: 62P10


accelerated failure time model; COVID-19; mortality analysis; survival analysis.


J. Harlan, Analisis Survival, Depok: Gunadarma, 2017.

R. Saikia and M. P. Barman, "A Review on Accelerated Failure Time Models," International Journal of Statistics and Systems, vol. 12, no. 2, pp. 311-322, 2017.

A. A. Sayampanathan, C. S. Heng, P. H. Pin, J. Pang, T. Y. Leong and V. J. Lee, "Infectivity of asymptomatic versus symptomatic COVID-19," Lancet, vol. 397, pp. 93-94, 2021.

B. K. Mishra, A. K. Keshri, Y. S. M. B. K. Rao, B. Mahato, S. Ayesha, B. P. Rukhaiyyar, D. K. Saini and A. K. Singh, "COVID-19 created chaos across the globe: Three novel quarantine epidemic models," Chaos, Solitons & Fractals, vol. 138, p. 109928, 2020.

S. T. P. COVID-19, "Peta Sebaran," 1 August 2021. [Online]. Available: [Accessed 1 August 2021].

H. Nisa, A. S. F. Utami and A. D. Oktaviani, "Risk Factors for Mortality Among COVID-19 Patients in Asia: A Literature Review," Jurnal Ilmu dan Teknologi Kesehatan, vol. 9, no. 2, pp. 139-152, 2022.

D. Raut, P. Shrivastav and S. Parwe, "Causes of death in COVID-19 patients: A literature review," International Journal of Research in Pharmaceutical Sciences, vol. 11, no. 1, pp. 1918-1924, 2021.

W. Sanusi, A. Alimuddin and S. Sukmawati, "Model Regresi Cox dan Aplikasinya dalam Menganalisis Ketahanan Hidup Pasien Penderita Diabetes Melitus di Rumah Sakit Bhayangkara Makassar," Journal of Mathematics, Computations, and, Statistics, vol. 1, no. 1, pp. 62-77, 2018.

E. A. Jal-Usman, A. A. Nads and D. G. Langamin, "Accelerated Failure Time (Aft) Model: Determining The Factors Associated in the Recovery of Patients Diagnosed with Covid-19," Advances and Applications in Statistics, vol. 79, pp. 1-9, 2022. DOI: 10.17654/0972361722056.

M. H. R. Khan and T. Howlader, "Stability Selection for Lasso, Ridge and Elastic Net Implemented with AFT Models, Running Title: Stability Selection for AFT Models," arXiv:1604.07311v1 [stat.ME] 25 Apr 2016, 2016.

K. Granville and Z. Fan, "Buckley-James Estimator of AFT Models with Auxiliary Covariates," PLoS ONE, vol. 9, no. 8, p. e104817, 2014. DOI: 10.1371/journal.pone.0104817.

M. Pang, R. W. Platt, T. Schuster and M. Abrahamowicz, "Flexible Extension of The Accelerated Failure Time Model to Account for Nonliniear and Time-Dependent Effects of Covariates on The Hazard," Statistical Methods in Medical Research, vol. 30, no. 11: 096228022110417, 2021. DOI: 10.1177/09622802211041759.

Rachmaniyah, Erna and Saleh, "Analisis Survival dengan Model Accelerated Failure Time Berdistribusi Log-normal," Universitas Hasanudin, 2014. Available at

M. J. Crowther, P. Royston and M. Clements, "A Flexible Parametric Accelerated Failure Time Model and The Extension to Time-Dependent Acceleration Factors," Biostatistics, vol. kxac009, 2022.

Sulantari and W. Hariadi, "Analisis Survival waktu sembuh pasien COVID-19 di Kabupaten Banyuwangi," Transformasi: Jurnal Pendidikan Matematika dan Matematika, vol. 4, no. 2, pp. 375-386, 2020.

A. F. Majeed, "Accelerated Failure Time Models: An Application in Insurance Attrition," The Journal of Risk Management and Insurance, vol. 24, no. 2, pp. 12-35, 2020.

N. Kumar, S.-A. Qi, L.-H. Kuan, W. Sun, J. Zhang and R. Greiner, "Learning Accurate Personalized Survival Models for Predicting Hospital Discharge and Mortality of COVID-19 Patients," Scientific Reports, 12, 4472, 2022.

J. P. Klein and M. L. Moeschberger, Survival Analysis Techniques for Censored and Truncated Data, Second Edition, New York: Springer, 2003.

N. I. Fitriani, "Tinjauan Pustaka COVID-19:Virologi, Patogenesis, dan Manifestasi Klinis," Medika Malahayati, vol. 4, no. 3, pp. 194-201, 2020.

D. F. Moore, Applied Survival Analysis Using R., Switzerland: Springer, 2016.

S. P. Khanal, V. Sreenivas and S. K. Acharya, "Accelerated Failure Time Models: An Application in the Survival of Acute Liver Failure Patients in India," International Journal of Science and Research (IJSR), vol. 3, no. 6, pp. 161-166, 2012.

G. Rodriguez, Generalized Linear Models, lecture notes,, 2022.

A. Faruk, "The Comparison of Proportional Hazards and Accelerated Failure Time Models in Analyzing the First Birth Interval Survival Data," Journal of Physics: Conference Series, vol. 974 012008, 2018.

Full Text: PDF

DOI: 10.15408/inprime.v4i2.25675


  • There are currently no refbacks.