The normalisation of resource efficiency measures in healthcare facilities: the case of energy and water

Amunjela, Abimelek Shikongo (2019-04)

Thesis (MEng)--Stellenbosch University, 2019.

Thesis

ENGLISH ABSTRACT: Energy and water consumption performance comparisons are used at a healthcare management and policy formulation level when formulating benchmarks and energy improvement targets. Normalising for the differences between hospitals is a key part of hospital consumption performance comparisons. It ensures that the measures used in these comparisons are commensurate, thereby increasing the reliability and robustness of the comparisons. Currently, the measures used in these comparisons are only normalised for the size of a hospital and are not adjusted to account for the inter-hospital differences in the level of medical service provision. This study investigated the feasibility of including normalisation factors that are representative of the level of medical service provision in the normalisation model used to compare the energy and water performance of hospitals. The complexity and level of specialisation of the composition of a hospital’s diagnostic caseload and the output of a hospital were used to quantify the level of medical service provision at a hospital. Measures were also formulated to quantify the size of a hospital. Statistically-based modelling methods were used to conduct an exhaustive analysis of the relationships between combinations of the normalising factors in the analysis and the energy and water consumption of the respective hospitals. Multiple linear regression (MLR) models were developed for all the possible combinations of normalising factors. These models were used to assess and rank the explanatory power provided by each combination of normalising factors in explaining the variations in the energy and water consumption of hospitals. Based on these MLR analyses and the rankings of the explanatory power provided by the respective models, it was concluded that the level of medical service provision of a hospital (as represented by its output, complexity and level of specialisation), does not significantly contribute to increasing the reliability or robustness of the current normalisation model. Furthermore, accounting for the level of medical service provision in the normalisation model would complicate the model without providing any significant additional explanatory power or increasing the objectivity of hospital consumption performance comparisons.

AFRIKAANSE OPSOMMING: Energie- en waterverbruik-prestasievergelykings word gebruik op 'n gesondheidsorgbestuurs- en beleidsformuleringsvlak wanneer maatstawwe en doelwitte vir energieverbetering geformuleer word. Normalisering vir die verskille tussen hospitale is 'n belangrike aspek wanneer hospitaalverbruikverrigting vergelyk word. Normalisering verseker dat die maatstawwe wat in hierdie vergelykings gebruik word regverdig is, en verhoog dus die betroubaarheid en robuustheid van die vergelykings. Tans word daar in die maatstawwe wat in hierdie vergelykings gebruik word slegs genormaliseer vir die grootte van 'n hospitaal, en word dit nie aangepas om byvoorbeeld die verskille tussen hospitale in die vlak van mediese diensverskaffing nie in berekening te bring nie. Hierdie studie het ondersoek ingestel na die uitvoerbaarheid van die insluiting van normaliseringfaktore wat verteenwoordigend is van die vlak van mediese diensverskaffing in die normalisasiemodel wat gebruik word om die energie- en waterprestasie van hospitale te vergelyk. Die kompleksiteit en vlak van spesialisering van die samestelling van 'n hospitaal se diagnostiese gevallelading en die uitset van 'n hospitaal is gebruik om die vlak van mediese diensverskaffing by 'n hospitaal te kwantifiseer. Maatstawwe is ook geformuleer om die grootte van 'n hospitaal te kwantifiseer. Statistiek-gebaseerde modelleringsmetodes is gebruik om 'n volledige ontleding van die verhoudings tussen kombinasies van die normaliseringfaktore in die analise en die energie- en waterverbruik van die onderskeie hospitale te doen. Meervoudige lineêre regressie (MLR) modelle is ontwikkel vir al die moontlike kombinasies van die normaliseringfaktore. Hierdie modelle is gebruik om die verklarende krag wat deur elke kombinasie van normaliseringfaktore verskaf word te assesseer en te rangskik om die variasies in die energie- en waterverbruik van hospitale te verduidelik. Op grond van hierdie MLR-ontledings en die ranglys van die verklarende krag wat deur die onderskeie modelle uitgewys word, is daar tot die gevolgtrekking gekom dat die vlak van mediese dienslewering van 'n hospitaal (soos mee gebring deur die lewering, kompleksiteit en spesialiseringsvlak) nie beduidend bydra tot die verbetering van die betroubaarheid of robuustheid van die huidige normaliseringsmodel nie. Om die vlak van mediese diensverskaffing in die normaliseringsmodel in ag te neem sou die model onnodig kompliseer sonder om enige beduidende addisionele verklarende krag te gee, of om die objektiwiteit van die vergelykings van hospitaal-verbruikverrigting te verhoog.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/106122
This item appears in the following collections: