Mandela University Statistics lecturer Dr Sisa Pazi’s research culminated in a statistical model for assessing disease severity at ICU admission, which could also be used to predict in-hospital mortality.
The predictive scoring system uses patient information, such as age, and several other variables to obtain a numerical value, or score, for each individual at ICU admission.
The higher the score, the more ill the patient; and this score is then used to estimate the risk of in-hospital mortality.
Dr Pazi was the statistical consultant for the research project, which began in 2017, spearheaded by Livingstone Hospital’s Adult Critical Care Unit head, Dr Elizabeth van der Merwe, and Mandela University Department of Statistics Professor Gary Sharp.
“Their plan was to collect a large database of patient information, which could be used for various interdisciplinary research projects.
Prof Sharp then approached me to be statistical consultant, with a view to collecting data for a doctoral study.
“This was part of a broader interdisciplinary research project which culminated in five research manuscripts, published in DHET-accredited journals, and six conference presentations.”
Pioneering research
Several papers have been published, but the umbrella topic, ‘Sustainable Critical Care: Biostatistics Empowers Life-Enhancing Decisions’, encompasses all the work done, says Dr Pazi, who also serves as a South African Statistical Association member, involved in facilitating bursaries and scholarships for undergraduate statistics students in South Africa.
His research paper, titled ‘Prediction of in-hospital mortality: an adaptive severity-of-illness score for a tertiary ICU in South Africa’, encapsulated the research needed for his doctoral study.
“This is the first attempt to develop a predictive scoring system based on South African data. The purpose of the system is to identify high-risk patients who can then be treated with the urgency required. This aligns with a patient-first approach.”
Livingstone ideal location
The choice of Livingstone for a research project of this nature was fortuitous, says Dr Pazi. “The timing aligned with Dr Van der Merwe’s – and others’ – need to broaden research, and the plan by our University to start a medical programme.
“Statistical skills in the health sciences are highly specialised, and this offered an opportunity to bridge that gap.”
Studies quickly showed how beneficial a predictive scoring system is, positively impacting both medical professional and patient.
“Predictive scoring systems are useful for standardising research and comparing the quality of ICU patient care. In addition, for resource-constrained populations such as the South African public health care sector, predictive scoring systems are useful in facilitating triage guidelines.”
The system was initially developed using patient information outside the African continent, and then tested at Livingstone Hospital.
“Although we found that it was adequate for use, it was clear that there was room for improvement. The current research paper then proposed a revised model based on data collected at Livingstone Hospital itself. This revised model can now predict in-hospital mortality with higher accuracy, allowing high-risk patients to be identified for urgent interventions.”
“I wanted to use maths to solve real-world problems”
Born in East London to a domestic worker, Sisa Pazi was raised by his grandmother, Mavis Ndaro, whom he describes as a “superwoman”.
Mathematics was Dr Pazi’s favourite subject at school, and he always knew that he wanted a career in numbers.
“I enrolled for a BSc at Nelson Mandela University in 2011. After learning of career opportunities in the field of statistics, I knew that as a statistician, I could use mathematical tools to solve real-world problems.”
Dr Pazi majored in statistics and applied mathematics, earning his BSc and MSc in 2015 and 2017 respectively, and graduating with his PhD in Mathematical Statistics in April this year.
Married to Sinoyolo and father to Zingce, 13, and one-year-old Lucwangco, he credits his mentors – grandmother Mavis Ndaro, sisters Noluthando Pazi and Unathi Faku, and Professor Gary Sharp – with putting him firmly on the path to finding solutions through statistics.