An analysis of the upstream supply chain for second-line drugs for multidrug-resistant tuberculosis

Lingervelder, Deon (2017-03)

Thesis (MEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Systematic problems in the supply chain of second-line anti-TB drugs (SLDs) for multidrug- resistant tuberculosis (MDR-TB) are well documented and contribute significantly to the comprehension of the difficulties preventing successful control of the disease. Though literature contains a wealth of proposed changes to global SLD supply chain policies, there is a significant research gap related to quantitative modelling of the SLD supply chain to accurately predict the expected impact of these proposed changes on the availability of SLDs. The global SLD supply chain consists of two components: (i) the ‘upstream’ component which includes all activities from the manufacturing of the active pharmaceutical ingredient through to the warehousing of drugs prior to shipment; and (ii) the ‘downstream’ component which includes in-country warehousing and delivery of drugs to various healthcare facilities. A prominent problem in the supply chain, is the erratic demand patterns, since these prohibit accurate forecasting and effective planning. Consequently, manufacturers are forced to produce drugs in inefficient batch sizes, causing higher prices and longer, inconsistent lead times. A possible solution to address this problem, is the implementation of a large buffer stockpile directed at (i) preventing stock-outs and treatment interruptions, and (ii) combining and timing orders to permit current manufacturers to produce medicines more efficiently. The aim of this study is to model a part of the upstream supply chain of MDR-TB SLDs and to evaluate the impact of implementing such a buffer stockpile. The supply chain is modelled using system dynamics and the model is used to evaluate the likely impact of a range of alternative inventory management policies on the supply chain performance. Three different SLD formulations are included in the model to ensure that the recommendations based on this research are robust. These formulations, namely capreomycin, kanamycin and cycloserine, account for approximately 58% of the total procurement costs of the current supply chain. The modelling results indicate that the inventory policies that will most likely lead to the most significant improvement in the supply chain performance, are the policies that implement a reorder quantity based on an exponential smoothing forecast of previous demand, specifically when a smoothing factor of either 0.1 or 0.5 and a high reorder point are implemented. This research contributes to the current academic literature by increasing the understanding of the upstream SLD supply chain, by providing a quantitative evaluation of the expected impact of suggested changes to the supply chain, and by presenting an example of an application of the system dynamics modelling approach that is not common in literature.

AFRIKAANSE OPSOMMING: Sistematiese probleme in die voorsieningsketting van tweede lyn anti-TB-middels (SLDs) vir multi-weerstandbiedende tuberkulose (MDR-TB) is goed gedokumenteer en maak ‘n aansienlike bydrae om die probleme wat die suksesvolle beheer van die siekte voorkom, te verstaan. Alhoewel literatuur 'n rykdom van voorgestelde wysigings aan die globale SLD voorsieningsketting bevat, is daar 'n beduidende navorsing gaping wat verband hou met die kwantitatiewe modellering van die SLD voorsieningsketting om die verwagte impak wat hierdie voorgestelde wysigings aan die beskikbaarheid van SLDs sal hê, akkuraat te voorspel. Die globale SLD voorsieningsketting bestaan uit twee komponente: (i) die 'stroomop’ komponent wat alle aktiwiteite van die vervaardiging van die aktiewe farmaseutiese bestanddeel, tot die verpakking van die teenmiddels vir distribusie, insluit; en (ii) die ‘stroomaf’ komponent wat die binnelandse (nasionale) pakhuise en die aflewering van teenmiddels na verskeie gesondheidsorg fasiliteite insluit. 'n Prominente probleem in die voorsieningsketting is die dinamiese en wisselvallige aanvraag patrone vir die teenmiddels, aangesien dit akkurate vooruitskatting en effektiewe beplanning verhoed. Gevolglik word vervaardigers gedwing om die teenmiddels in onekonomiese hoeveelhede te vervaardig wat hoër pryse en langer wagtye tot gevolg het. 'n Moontlike oplossing om hierdie probleem aan te spreek, is die implementering van 'n groot buffer voorraad wat gerig is op: (i) die voorkoming van onderbrekings in die behandeling van MDR-TB, deur te verseker dat daar altyd teenmiddels op voorraad is; asook (ii) om beter tydsberekening toe te laat om die verskeie bestellings te kombineer, sodat huidige vervaardigers die teenmiddels meer doeltreffend kan produseer. Die doel van hierdie navorsing is om 'n deel van die ‘stroomop’ voorsieningsketting van MDR-TB SLDs te modelleer en om die impak wat 'n buffer voorraad op die voorsieningsketting kan hê, te evalueer. Die model is ontwikkel deur ‘n Stelsel Dinamika Modellering benadering en word gebruik om die moontlike impak van 'n verskeidenheid alternatiewe voorraadbestuur beleide op die voorsieningsketting te evalueer. Drie verskillende SLD formulerings is ingesluit in die model om te verseker dat die aanbevelings, op grond van hierdie navorsing, robuus is. Hierdie formulerings, naamlik capreomycin, kanamycin en cycloserine, is verantwoordelik vir ongeveer 58% van die totale aankoopkoste van die huidige voorsieningsketting. Die resultate van die modellering dui daarop dat die voorraadbestuur beleid wat meer waarskynlik sal lei tot die mees beduidende verbetering in die voorsieningsketting, is die beleide waar die herbestel hoeveelheid bepaal word op grond van 'n eksponensiële ‘gladstryking’ voorspelling van vorige aanvraag, spesifiek wanneer ‘n ‘gladstryking’ faktor van óf 0.1 of 0.5 en 'n hoë herbestelvlak geïmplementeer word. Hierdie navorsing dra by tot die huidige akademiese literatuur deur die begrip van die stroomop SLD voorsieningsketting te verbeter, deur 'n kwantitatiewe evaluering van die verwagte impak van voorgestelde wysigings aan die voorsieningsketting te verskaf en deur 'n voorbeeld van ‘n Stelsel Dinamika Modellering te verskaf, wat nie algemeen in die literatuur is nie.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/101044
This item appears in the following collections: