Doctoral Theses

Permanent URI for this collection

http://192.168.18.118:4000/handle/20.500.12661/401

Browse

Browsing Doctoral Theses by Author "Ismail, Abbas"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Statistical models and quality indicators for evaluation of clinical care among pediatric patients living with HIV/AIDS in Dar es salaam, Tanzania
    (The University of Dodoma, 2021) Ismail, Abbas
    Pediatric HIV patients are not randomly distributed and health services are not randomly allocated, and available methods for measuring Tanzania’s health care have not adequately addressed methodological strength of the indicators on providing valid estimates. This study intended to identify statistical methods and quality indicators for pediatric health care evaluation in Tanzania. The quality indicators were computed using proportions. The cut off values of the quality indicators in classifying levels of HIV pediatric health services were obtained. Inferential analysis was involved in obtaining quality indicators with clinical relevance and to achieve, different forms of survival analysis and Generalized Estimating Equations were fitted. After identifying quality indicators with clinical significance, ranking of health facilities based on risk adjusted models was also done. The indicators ‘eligart’ and ‘eligcotrm’ have shown clinical relevance on performance measures of facilities. Independence and exchangeable count generalized estimating equations, and frailty models for mortality rates were preferred methods in evaluating clinical relevance of the process quality indicators. The study concluded that process indicators of providing early ART initiation and provision of cotrimoxazole prophylaxis were important functions in monitoring and evaluation of facility performance. Hierarchical models for facility clustered level data can be used for performance assessment. Further studies should define more indicators and evaluate whether or not the inclusion of more detailed patient and facility health characteristics in adjusted models significantly alter quality reporting system.