Welcome to Dr Sandro Tsang’s Homepage

I was awarded a PhD in Empirical Economics, and have worked in academic institutes of several countries. Since 2012, I have been a tutor on a fully online master’s degree programme in Public Health. The programme was developed by Peoples-uni, and the award accredited initially by Manchester Metropolitan University in the UK and subsequently by the EUCLID University. I also work with NextGenU.org to revise public health courses delivered to graduate students of medical schools. This experience has stimulated me to learn medical research methodologies applicable to evidence-based clinical practice and health policy-making.

Clinical reasoning is my primary research interest. The understanding of clinical behaviours critical for deriving operationalisable healthcare policies is a core interest, but this topic is scarcely covered in economics textbooks. I derived a mathematical model that captures the uncertainties involved in prescribing antibiotics and the consequences as a 3-month MSc dissertation (here is a published version). I subsequently researched into other clinical decision problems, including eHealth, influenza-like illnesses, etc. My PhD thesis explored impacts of implementing eHealth to facilitate clinical reasoning from the perspectives of epistemology and practising humanistic medicine. In a wish to further develop this eHealth model, I was lucky to reach Professor Michael Loughlin who was prominent in the philosophy of humanistic medicine and its related research. My ideas inspired him to guide me to take the research further. Without his guidance, I might not have been motivated to write up an interdisciplinary theory which led to the award of an ESPCH Essay Prize (reputed researchers were also in the competitor pool). This theory portrays the Arrow Physician to be an ideal medical practitioner in modern health systems. S/he is required to deliver patient-centred care in a humanistic approach, and to reason like an economist at the same time. I postulate that establishing polycentric health systems is crucial to substantiate her/his mission, and to improve healthcare resource allocations through collective ethical actions. My research also overlaps with policy-making for infectious disease surveillance and mitigating antimicrobial resistance. It also involves turning my self-taught skills in programming (including R, python, SQL, bash, etc.) into action.

I mostly apply programming skills to analytic or typesetting levels. However, I have spotted room for improving R packages koRpus, mice and miceadds. I do practise the art of programming – solving problems efficiently with basic and minimal commands. This practice is shown in the codes stored in my GitHub repos. I have only presented some R and python codes there, although this webpage demonstrates applications of four computational languages.

I am active on LinkedIn and X (twitter), where I share various academic topics and some computation tricks. I welcome your contact through email or filling in the form. I hope the (zoomable) slideshow on the sidebar entertains you too. If you are visiting this webpage with a smartphone, please rotate it to see the slideshow.

Publication list

The embeded pdf file may not be loaded instantly, but you may open it in a new tab.

Code Samples

My GitHub repos present a small sample of R and python codes. The X-ray-Pytools include tools for quality check, editing,... building convolutional neural networks from X-ray images. I was a novice of computer vision and developed the full-stack skills for handing image data for a 3-month project.

The scrapping scripts were written to serve my curiosity. I find that scrapping Google Scholar (GS) is increasingly applied to conduct systematic literature review. The researchers either scrapped the first 100 abstracts or they never explained how they scrapped all the abstracts.

The first approach does not fulfill an exhaustive search criterion of systematic review. We know that the robots.txt file of GS prohibits us from scrapping more than 100 abstracts. The second approach can be implemented by scrapping the downloaded webpages. I created a timer to download the webpages in random time intervals. I have never implemented this approach in my own research, because I am not wealthy enough to enter into a legal battle with the conglomerate. I wrote up the scripts to simply verify the claim. They may be excuted for other journal databases. Scrapping GS violates its user policies.

N.B. If you are a smartphone user, the form will prompt you to the first page. Please accept my apologies. I will resolve it after fixing some urgent matters.