Introduction

Liver disease is currently the 5th biggest cause of mortality in England and Wales. The

UK liver disease crisis has been captured and extensively analysed by the Lancet

commission group in 2014 (and subsequent versions). This landmark publication has

produced a blueprint for addressing the burden of liver disease in the UK. The scope

of the report does not only cover common liver diseases, but also rarer causes of

hepatic pathology often called orphan liver diseases. It is estimated that there are 54

million people living with a rare disease in Europe and North America. There is also a

wider call to consider integrated care for all patients with liver disease though

effective chronic disease management.

In this thesis, I argue that rare diseases are also chronic diseases and should be viewed

through the prism of the chronic care model (CCM) which has been successfully used

historically for commoner conditions such as diabetes and chronic obstructive

pulmonary disease (COPD). The low prevalence of rare liver diseases leads to paucity

of data both from clinical trials as well as the real world. The European Union

Committee of Experts on Rare Diseases (EUCERD) was set up with the purpose of

encouraging the exchange of relevant experience, policies, and practices in rare

diseases among member states. It became the prelude to the European Reference

Networks (ERNs) which were set up at a later stage to underpin the provision of robust

governance and policy in data collection and registration in rare diseases within the

European Union (EU). Despite this framework, a comprehensive blueprint of how to

create an effective and contemporary registry for rare liver diseases does not exist to

date, despite there being approximately 20 non-cancer rare hepatic conditions.

In this project, I used primary biliary cholangitis (PBC) as an example of a rare liver

disease. My aim was to initially understand whether there is much evidence in the

literature on the use of CCM for rare liver diseases, and subsequently, develop a model

for creating registries for rare liver disease. Using this model, I next set out to build a

regional registry for PBC using real world regional data in the county of Surrey, UK.

Thereafter, I sought to use the registry to examine whether it had yielded meaningful

clinical outputs for patients with PBC in the county of Surrey.

Methods

Before I set up the regional registry for PBC in Surrey, I sought to identify European

non-cancer registries for patients with rare liver diseases. Using identified literature

and data from those registries, I was able to develop my own model for an aspirational

data registry for rare liver diseases. I used this model as a theoretical cornerstone to

build the PBC registry, which also allowed data linkage from primary, secondary, and

tertiary care. Following the necessary applications and approvals from the Health

Research Authority, data were collected from primary, secondary, and tertiary care.

For the primary care data, I used the database of the Research and Surveillance Centre

(RSC) of the Royal College of General Practitioners (RCGP), which holds data on more

than 1.2 million patients in England. I focused on GP surgeries that have consented to

data collection in the county of Surrey. Relevant data were captured for the registry.

Moreover, I also used the primary care data to explore whether I could develop a

novel ontology to search for patients with PBC in primary care datasets. I collected

real-world secondary care data from three regional NHS hospital Trusts including

Royal Surrey NHS Foundation Trust (RSNHSFT), Ashford and St Peter’s hospital (ASPH)

and East Surrey hospital (ESH). Following the necessary application, tertiary centre

data were obtained from patients in the county of Surrey who had received a diagnosis

of PBC. Once the data were mined and uploaded onto the registry, they were analysed

using Statistical Package for the Social Sciences (SPSS) v28.

Results

My initial literature review using a systematic approach identified that there was

limited use of the chronic care model in patients with rare liver disease. There were

only 6, 11, 1, 13, 2 and 0 studies discussing individual components of the CCM for

Autoimmune Hepatitis (AIH), PBC, Primary Sclerosing Cholangitis (PSC), Wilsons

disease (WD), Alpha-1 Antitrypsin Deficiency (A1AD) and Lysosomal Acid Lipase

deficiency (LALd) respectively. I did not identify any studies using the full CCM for any

hepatic orphan liver diseases. One of the components of the CCM is the use of clinical

information systems and registries. There was very little identified literature on the

use of disease registries for rare liver disease. A separate literature review was carried

out to appreciate the cardinal components and crucial elements of established

registries for rare liver disease. The review identified 37 European registries, which

were analysed and led to the development of a novel registry design blueprint. Using

information from the design of these registries I developed a blueprint for the

development of a patient registry for rare liver diseases consisting of 9 stages under

three phases: the theoretical, technical and maintenance phases. I used this model to

build a regional PBC registry.

I searched 218,099 primary care records and developed a novel clinical ontology to

identify patients with PBC. Using this ontology, I identified 58 patients with likely PBC

and 2317 patients with probable PBC. There were 32 new cases of PBC. These data

were linked to secondary care data and in total, the registry held regional real-world

data for 403 patients with PBC. The ratio of male: female was found to be

approximately 1:10 and the average age at diagnosis was 59. Most patients were

White Caucasian (93%). Fatigue, itching, and arthralgia were the commonest

presenting symptoms. Using reference criteria to assess response, I found higher

response rates in underdosed patients. Similarly, underdosing patients appears to also

yield better response rates when the Barcelona, Paris-2 and Ehim criteria were used

to assess response. On the contrary, overdosing patients confers better response

rates when using the Paris-I, Toronto, Rotterdam and Momah Lindon criteria.

Discussion

It is believed that rare diseases may affect as much as 6-8% of the EU population across

its 28 member states, and yet there is little published consideration for these

conditions to be viewed through models of chronic care. One of the novel outputs of

this work is the development of a toolkit for designing registries for rare liver disease

(and not only). Therefore, this work complements the efforts of loco-regional,

national, and international groups seeking to establish robust systems for data

collection and analysis for orphan liver diseases. Another novel output of this thesis is

the development of a PBC ontology, which was used to search primary care records.

Using this tool, I was able to identify (i) established cases of PBC not known to

local/regional secondary care providers and (ii) de novo PBC cases, not previously

identified either in primary or in secondary care. There are many PBC probable cases

whose data merit further careful evaluation, and it is possible that many of these cases

are true PBC cases. The Surrey PBC registry is probably the largest real-world PBC

database, which I have utilised to describe in detail, demographics, geoepidemiology,

referral timelines, clinical management, pharmacotherapy, natural history of disease,

and survival outcomes of patient with PBC.