Impact of the COVID-19 Crisis on Treatments, Natural History, Health Care Resource Utilisation, and Medical Cost in Hypertrophic Cardiomyopathy in England

Date of Approval: 
2021-04-09 00:00:00
Lay Summary: 
Hypertrophic cardiomyopathy (HCM) is a primary myocardial (i.e., heart muscle) disorder defined by thickening (i.e., hypertrophy) of the left ventricular (LV) wall that is unexplained by secondary causes. HCM, particularly obstructive HCM (oHCM) occurring when the LV wall thickens such that blood flow to the aorta is reduced/blocked, is associated with increased risks of heart failure (HF), atrial fibrillation/flutter, and sudden cardiac death, among other outcomes. Evidence from the US portends substantial unmet need and disease burden in HCM, which is likely exacerbated by the ongoing COVID-19 crisis given impacts on the accessibility of essential cardiology services. The current study proposes to use linked Clinical Practice Research Datalink (CPRD) / Hospital Episode Statistics (HES) data to study the impact of the COVID-19 crisis on treatments, natural history, health care resource utilisation, and medical cost of HCM in England. Specifically, we first seek to understand the epidemiology, patient characteristics, treatment history and treatment patterns, natural history, health care resource utilisation, and medical cost of HCM patients in England, and then more importantly, how treatment use and treatment patterns, natural history, health care resource utilisation, and medical cost changed after the initiation of COVID-19 restrictions in England, which are assumed to have begun in earnest 16th March, 2020. The public health benefit of this study is highlighted by the lack of comprehensive understanding of HCM in England and how the COVID-19 crisis has impacted treatments, natural history, health care resource utilisation, and medical cost in HCM in England.
Technical Summary: 
With the linked CPRD-HES database, patients with HCM and its subtypes (oHCM, symptomatic oHCM, asymptomatic oHCM, and non-obstructive HCM [nHCM]) will be selected using validated codes indicative of diagnoses, symptoms, and procedures. A control cohort of patients with no HCM will be matched with a 3:1 ratio to each case on age, sex, region, and index year. It is hypothesised that compared with controls, HCM patients had increased risks of developing clinical events and increased health care resource utilisation and medical cost. Additionally, it is hypothesised that the initiation of COVID-19 restrictions in England caused an interruption in HCM treatments, and thereby increased incidences of clinical events, and decreased health care resource utilisation and medical costs, potentially with a larger impact on general practitioner (GP) visits than inpatient admissions. In the analysis, first, the frequency of HCM and its subtypes will be calculated as a proportion of all patients in the database. Second, treatment history and treatment patterns in HCM will be summarized and compared before vs. after the implementation of the COVID-19 restrictions in England. Third, natural history will be summarized as incidence rates of and time to clinical conditions and events. This will be compared between HCM patients and controls and before vs. after the COVID-19 restrictions. Fourth, health care resource utilisation will be evaluated as the proportions and incidence rates of patients using each type of resource in primary and secondary care. Medical costs will be assessed using unit costs for each type of health care resource utilisation obtained from National Health Service (NHS) reference costs, and applied to the number of visits. Health care resource utilisation and medical costs will be compared between HCM patients vs. controls and before vs. after the COVID-19 restrictions using generalised estimating equation (GEE) models.
Health Outcomes to be Measured: 
Epidemiology: Proportions of patients in the linked CPRD-HES database with HCM, oHCM, symptomatic oHCM, asymptomatic oHCM, and nHCM in England.; Treatment history and treatment pattern: Treatment history variables include time to initial treatment after the initial diagnosis, treatment sequence, treatment use (i.e., number and proportion of patients using each class of treatment for each line of therapy), average treatment duration for each line of therapy, and gap time between adjacent lines of therapy. Treatment patterns include the persistence and adherence associated with existing pharmacologic treatments. Lines of treatments will be constructed at the treatment class level (i.e., beta blocker [BB], calcium channel blocker [CCB], disopyramide, combination therapy, SRT [septal myectomy, alcohol septal ablation], cardiac resynchronization therapy [CRT], implantable cardioverter-defibrillator [ICD] insertion, and other procedures [i.e., heart transplant and pacemaker without bradyarrhythmia]). Adherence to pharmacotherapies will be summarized as the proportion of days covered (PDC). Natural history: The incidence rate of and time to the following clinical conditions will be assessed: acute renal failure, atrial fibrillation/flutter, chest pain, HF, cardiac arrest, stroke, supraventricular tachycardia, ventricular tachycardia, syncope, ventricular fibrillation, and indication of symptomatic oHCM (for patients with asymptomatic diagnosis only). In addition, the following clinical events will be assessed: any hospitalisation, any emergency room (ER) visit, heart transplantation, CRT, ICD placement, SRT, death, and cardiovascular-disease-related (CVD) death.; Health care resource utilisation: Health care resource utilisation outcomes will include the proportions and the incidence rates of primary care contacts (general practice consultations, telephone consultations, home visits, emergency visits, and out-of-hours consultations) and secondary care contacts (hospital inpatient admissions [including length of stay], outpatient visits, day case visits, emergency visits). Specialty visits, blood tests, imaging, and procedures relevant to HCM may also be considered. All-cause health care resource utilisation will be studied. CVD-related resource utilisation may also be studied. Medical cost: Medical cost outcomes will include mean costs associated with each type of health care resource utilisation, using unit costs for each type of health care resource utilisation obtained from the NHS reference costs and the Personal Social Services Research Unit. All-cause medical costs will be studied. CVD-related medical costs may also be studied.
Application Number: 
21_000342
Collaborators: 

Carla Zema - Chief Investigator - MyoKardia

Riley Taiji - Corresponding Applicant - Analysis Group, Inc.

Jenny Zhou - Collaborator - Analysis Group, Inc.

Jipan Xie - Collaborator - Analysis Group, Inc.

Sophia Li - Collaborator - MyoKardia

Yan Wang - Collaborator - Analysis Group, Inc.

Linkages: 
HES Accident & Emergency, HES Admitted Patient Care, HES Diagnostic Imaging Dataset, HES Outpatient, ONS Death Registration Data