Retrospective analysis of the burden of vaso–occlusive events experienced by sickle cell disease (SCD) patients
Healthy red blood cells are shaped like a disc with indents either side. Their role is to carry oxygen throughout the body. When a person has an inherited condition called sickle cell disease (SCD), their red blood cells become crescent, or “sickle”, shaped. These sickle cells die earlier than normal red blood cells, leading to a constant shortage of red blood cells. Additionally, the shape of these cells means they can get stuck in blood vessels and clog blood flow. When blood vessels become blocked to the point where tissues are deprived of oxygen, this is called a vaso-occlusive crisis (VOC). These events are extremely painful and can lead to further health complications if left untreated.
SCD is estimated to affect 14,000 people in the UK and is particularly common in people with an African or Caribbean family background. Little research has comprehensively characterised the frequency of VOCs in patients with SCD and documented their healthcare service usage.
Novartis, a pharmaceutical company, collaborated with ICHP to quantify the potential cost savings the NHS could realise with a treatment such as crizanlizumab, a drug patented by Novartis, which aims to prevent VOCs associated with SCD.
This study had two main objectives. The first was to quantify the frequency of VOCs in patients with SCD over 16 years of age. The second objective was to quantify healthcare resource utilisation of this group across North West London’s healthcare system.
We extracted data from the Discover database. Discover is a de-identified database of linked healthcare records for 2.8 million people registered to a GP practice in North West London (NWL). We identified all patients over 16 years of age with a recorded diagnosis of SCD and extracted data on VOCs and healthcare service use from 1st April 2015 until 31st March 2020. We included both VOC events that were explicitly coded and suspected VOC events that were not explicitly coded. We divided SCD patients into three groups: patients with no VOCs, patients with one VOC (low frequency group), and patients with two or more VOCs (high frequency group). In each group, we identified comorbidities and healthcare resource utilisation, as measured by the total number of interactions with primary care, inpatient, outpatient, A&E, community, and mental health care services. The costs of these services were extracted from the locally defined costs in the Discover database.
We identified 1,007 patients with SCD aged 16 years and older. This means the prevalence of SCD in NWL was 432 cases per million population, which is higher than the UK prevalence of 265 cases per million. Of these patients, 715 did not have any VOCs recorded across the 5 years we analysed, 67 had one VOC, and 225 had two or more VOCs. We found that 53% of VOCs were not explicitly recorded as a VOC but are suspected to be VOC events.
The cohort had 62,339 interactions with the healthcare system over the 5-year period. The average number of interactions per patient per year was highest for the high frequency VOC group at 27 interactions, followed by an average of 15 interactions for the low frequency VOC group, and 8 interactions for the no VOC group. All three groups most frequently interacted with outpatient services, followed by inpatient services for the high frequency VOC group, community services for the low frequency VOC group, and GP services for the no VOC group. The average cost per year of treating patients was £38,024, £12,780, and £4,170 in the high frequency, low frequency, and no VOC groups, respectively.
Our study provides confirmation that NWL has a high prevalence of SCD as compared to the rest of the UK. Most patients did not have a VOC during the five-year observation period. Over half of VOCs were not explicitly coded, which might indicate the healthcare system could improve the accuracy of how it records these events. Patients in the high frequency VOC group cost the healthcare system an average of £38,024 per year, which is approximately three times the cost of the low frequency VOC group, and nine times more costly per annum than the no VOC group. This indicates that significant reductions in cost and improvements in patient welfare could be gained from proactive identification and treatment of VOCs in patients with SCD.