Universal Journal of Public Health Vol. 10(6), pp. 555 - 562
DOI: 10.13189/ujph.2022.100601
Reprint (PDF) (489Kb)

Statistical Methods of Handling Ordinal Longitudinal Responses with Intermittent Missing Data

Aluko O. ^1,*, Mwambi H. ^2
1 Department of Biostatistics, Faculty of Health Sciences, School of Biomedical Sciences, University of the Free State, South Africa
² School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Private Bag X01 Scottsville 3209, Pietermaritzburg, South Africa

ABSTRACT

The rate of survival of human immunodeficiency virus (HIV) positive individuals resume to ameliorate with the usage of highly active antiretroviral therapy (HAART), yet pulmonary disease prevalence is growing unstopped amongst some of them. Handling missing data was a difficult challenge in the health data concept. We compared the effect of marginal to the proposed ordinal negative binomial models in handling intermittent missing observations for better numerical performance. The data used was characterised by monotone missing observations due to patients’ failure to declare their pulmonary conditions (lung infections and complications) and other vital health information. The use of multiple imputations is one of the latest techniques for handling missing observations, and this technique is relevant to different missing data mechanism assumptions, but occasionally profiles with the complication of the kind of parameters to be imputed and the mechanism underlying the incomplete data. This study focuses on the importance and application of the methods of handling missing health data. The proposed ordinal negative binomial model performs greatly than other models in adjusting for monotone missing data without imputation. In a real application, the proposed ordinal negative binomial model produces low estimates as against direct likelihood, mixed effects proportional odd, and MI-GEE models.

KEYWORDS
Ordinal Outcomes, Ordinal Negative Binomial, Multiple Imputations, Monotone Missingness

Cite This Paper in IEEE or APA Citation Styles
(a). IEEE Format:
[1] Aluko O. , Mwambi H. , "Statistical Methods of Handling Ordinal Longitudinal Responses with Intermittent Missing Data," Universal Journal of Public Health, Vol. 10, No. 6, pp. 555 - 562, 2022. DOI: 10.13189/ujph.2022.100601.

(b). APA Format:
Aluko O. , Mwambi H. (2022). Statistical Methods of Handling Ordinal Longitudinal Responses with Intermittent Missing Data. Universal Journal of Public Health, 10(6), 555 - 562. DOI: 10.13189/ujph.2022.100601.