This project is in the field of pharmacoepidemiology, public health and medical statistics. The aim is to examine the long-term effect of SARS-CoV-2 infection (including but not limited to long COVID) in mental health, especially psychiatric disorders and related medication use in Scotland population. Statistical methodologies and different epidemiological study designs will be explored to compare the robustness and reduce potential biases.

Clinical background

The continuing spread of coronavirus disease 2019 (COVID-19) through infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remains a major public health concern. Recent WHO COVID-19 statistics showed that there were cumulatively more than 776.8 million cases and 7.07 million deaths registered worldwide as of 10 November 2024 and the number is still increasing. Post COVID-19 Condition, commonly known as long COVID, is defined as the continuation or development of new symptoms 3 months after the initial SARS-CoV-2 infection and these symptoms can last for at least 2 months [1]. It was estimated that long COVID affects more than 10% of patients infected with SARS-CoV-2, with the most common symptoms including shortness of breath, fatigue, chest pain and cognitive dysfunction [2]. 

Studies based on electronic health records (EHR) and registry data have reported an increased risk of neurological and psychiatric disorders both in patients admitted to hospital, and individuals with mild or asymptomatic disease during 3-12 months after SARS-CoV-2 infection[3-7]. Researchers also found that vaccination before infection partly alleviates the risk of psychiatric sequelae up to 6-months’ follow up [8, 9]. A recent study using UK Biobank [9] showed that SARS-CoV-2 infected participants had higher risk of psychiatric disorders and psychotropic prescriptions, and risks were higher for hospitalised individuals than those not hospitalised and were lower for fully vaccinated people compared to those who were partially vaccinated or non-vaccinated. However, evidence on this topic seems limited in Scotland. 

Statistical Background

Statistical analyses of routine healthcare data are increasingly common for the assessment of the ‘real-world’ effect of medications, vaccination of infectious diseases.  Such studies are extremely powerful, as generally the sample size is very large, but have the common difficulties of observational studies – model associations only, an inability to attribute causality and biases, such as those associated with ascertainment, indication, healthy vaccinee and healthcare seeking behaviour.  There are many common study designs and analytical techniques – cohort, nested case control, target trial, regression discontinuity, propensity score matching and weighting, g-estimation and marginal structural models – and this project will systematically compare their use in the analysis of routine prescribing data.  We use mental health conditions (e.g. psychiatric disorders and psychotropic prescriptions) following from mild or severe SARS-CoV-2 infection in Scotland as the case study.
For this funded 3-year PhD, the student will conduct the following potential studies using data including Scottish prescribing data available in Public Health Scotland (PHS), Scottish Morbidity Records (SMR), and potentially British Heart Foundation Data Science Centre (BHF DSC).

• a systematic review (with meta-analysis) of current findings on the association between Covid-19 infection and mental health (UK and worldwide)
• detailed investigations of epidemiological study designs and methodology comparison using simulated data based upon Scottish prescribing data
• estimating risk of psychiatric disorder and psychotropic medication use using real-world data 
• analysis of COVID-19 data and follow-up in Scotland: methodologies to address biases 
• AI/machine learning based prediction models of psychiatric disorder events after COVID (optional)   

The position will be based at University of Strathclyde in Glasgow. You will closely work with the researchers in The Health Statistics research group from the Department of Mathematics and Statistics, and the Pharmacoepidemiology and Healthcare Research group from SIBPS at Strathclyde. Training in statistics and pharmacoepidemiology will be provided. You will be encouraged to attend local and overseas conferences and become part of international networks. This PhD training could lead to further research career opportunities in academia, health sciences industry such as contract research organisation (CRO), Pharmaceutical companies, and government regulatory.