An estimate of the cumulative paediatric prevalence of rare diseases in Ireland and comment on the literature

Most rare diseases (RDs) are complex, chronic, disabling and frequently life-threatening. Individuals with RDs, their families and caregivers often face difficulties accessing services, which can negatively impact health and all aspects of life such as social inclusion and access to education and employment [1].

Accurate estimates of any epidemiological indicators of RDs in the Republic of Ireland (ROI) are impeded by a number of factors such as: inadequate coding, lack of unique patient identifier and disjointed datasets. The ensuing health information gaps re-enforce the myth that RDs are uncommon and contribute to an inability to appropriately allocate resources, thus compounding health neglect for RD patients.

We sought to estimate the national cumulative paediatric incidence of RDs in a cohort of children born in the ROI in the year 2000. To maximise ascertainment, ORPHAcodes were manually assigned from narrative records along with ICD-10 coding, in line with our previous study [2]. Whilst this study was Dublin-centric, we are confident in ascertainment as 71% of the cohort were from outside the Dublin metropolitan area in agreement with governmental data (76%) [3] (Materials and methods are available as Supplementary Material).

Using the complete ORPHAcodes dataset, recording all RDs observed, our study identified 2283 out of a total of 54,789 livebirths born in the year 2000 in the ROI, diagnosed with a RD (4.2%) by 17 years, with 11.8% (270/2283) having died during this period.

The 4.2% of RD cases identified, analysed by Orphanet classification (Table 1), comprised 44.4% rare developmental defects during embryogenesis, 14.4% neurologic, 5.5% neoplastic and 1% rare infections. The results of this study show the ROI’s percentage of rare developmental defects during embryogenesis are greater than those recorded in other regions [4]. This discrepancy may be accounted for by the lack of termination of pregnancy legislation prior to 2019 and cultural norms at the time.

Table 1.

Categorisation of year 2000 cohort presenting by 17 years of age by rare disease category.

Rare disease category Rare disease	Number of cases (n)	Total [%]
Rare development defect during embryogenesis		1013 [44.4%]
Trisomy 21	(110)
Interauricular(atrial) communication	(43)
Ventricular Septal defect	(40)
Intellectual disability	(39)
Cleft palate	(34)
Transposition of the great arteries	(30)
Aorta coarctation	(30)
Neurofibromatosis	(20)
Spina bifida	(26)
Partial autosomal monosomy	(18)
Tetralogy of Fallot	(21)
Congenital diaphragmatic hernia	(19)
Other rare development defect during embryogenesis	(583)
Rare neurologic disease		329 [14.4%]
Epilepsy	(115)
Cerebral Palsy	(45)
Pervasive development disorder	(26)
Other rare neurological disease	(143)
Rare endocrine disease		136 [6.0%]
Precocious puberty	(48)
Growth disease	(32)
Congenital hypothyroidism	(13)
Other rare endocrine disease	(42)
Rare neoplastic disease		125 [5.5%]
Acute lymphoblastic leukaemia	(17)
Lymphoma	(9)
Acute myeloid leukaemia	(8)
Other rare neoplastic disease	(91)
Rare haematologic disease		102 [4.5%]
Sickle cell anaemia	(27)
Immune thrombocytopenic purpura	(13)
Typical haemolytic-uraemic syndrome	(13)
Other haematologic disease	(49)
Rare systemic or rheumatologic disease		101 [4.7%]
Immunoglobulin A vasculitis	(19)
Ehlers-Danlos syndrome	(19)
Kawasaki disease	(16)
Other systematic or rheumatologic disease	(47)
Rare skin disease		97 [4.3%]
Rare naevus	(26)
Congenital melanocytic naevus	(20)
Pilomatrixoma	(17)
Other rare skin disease	(34)
Rare bone disease		70 [3.1%]
Legg-Calve-Perthes	(13)
Osteogenesis imperfecta	(9)
Osteonecrosis	(8)
Other rare bone disease	(40)
Rare cardiac disease		55 [2.4%]
Non-genetic cardiac rhythm disease	(26)
Cardiomyopathy	(15)
Long QT syndrome	(6)
Other rare cardiac disease	(8)
Rare otorhinolaryngologic		47 [2.1%]
Non-syndromic genetic deafness	(29)
Congenital subglottic stenosis	(17)
Other rare otorhinolaryngologic disease	(1)
Rare respiratory disease		46 [2.0%]
Cystic fibrosis	(36)
Other rare respiratory disease	(10)
Rare inborn errors of metabolism		44 [1.9%]
Phenylketonuria	(7)
Disorder of porphyrin and haem metabolism	(5)
Other rare inborn errors of metabolism	(32)
Rare eye disease		37 [1.6%]
Non-syndromic congenital cataract	(11)
Other rare eye disease	(26)
Rare gastroenterological disease		22 [1.0%]
Hirchsprung’s disease	(10)
Undetermined colitis	(5)
Other rare gastroenterological disease	(7)
Rare infectious disease		23 [1.0%]
Meningococcal meningitis	(14)
Other rare infectious disease	(9)
Rare hepatic disease		6 [0.3%]
Chronic autoimmune hepatitis	(5)
Other rare hepatic disease	(1)

Categories where case numbers are <5 have been excluded to avoid disclosure issues.

Notably, 5.4% (n = 123) of the RD patients in the cohort had a RD known to be inherited in an autosomal recessive (AR) manner, and 1.4% (n = 31) had a RD confirmed to have X-linked inheritance, including de novo occurrences. A further n = 82 (3.6%) cases had RDs known to be genetic with heterogeneous modes of inheritance with a proportion due to AR and X-linked inheritance, but inheritance assignment in these cases was not possible. The only estimate in a comparable population is the DDD project in the United Kingdom (UK) [5] which estimated an incidence of 3.6% overall for AR disorders. The DDD study identified that between 6% of males and 6.9% of females had an underlying X-linked disorder [6], the majority of which occur de novo. The rate in our cohort is likely similar, but our inability to categorise cases is, in part, due to the testing methodology at the time this cohort was investigated.

We calculated a total mortality rate of 0.85% for the year 2000 cohort, in line with published Irish National Statistics Office (CSO) statistics [7], confirming the credibility of ascertainment in our study. Mortality in RD patients in this cohort was 11.8% by age 18. In line with previous research, the majority of RD deaths were in the first year of life (n = 165, 61%) [2].

Notably, 2013 of the 2283 RD patients identified in this study were alive at age 18. We estimate that a significant proportion of the 2013 have chronic conditions that required transition to adult care. From our analysis, 14.9% of total RD bed day usage was in the later teenage years (which represent 22% of years of the study). We noted ongoing, high bed day usage for teens with cystic fibrosis and sickle cell anaemia, while RDs with a similar prevalence, such as phenylketonuria, which have an ambulatory model of care, had no in-patient bed days.

It is recognised that our cohort study is a snapshot in time, and should this study be replicated, there would likely be variance, in both birth and death rates of different RDs. There have been improvements in survival in many paediatric disorders since the year 2000, such as development of new therapies for cystic fibrosis, spinal muscular atrophy, and new surgical treatment modalities. In addition, since the introduction of termination of pregnancy legislation, there will likely be a decrease in the rate of babies born with some congenital anomalies. The introduction of non-invasive prenatal testing screening is likely to reduce the incidence of all trisomies as some couples may elect for termination. In contrast, couples who avail of antenatal testing for cystic fibrosis and spinal muscular atrophy may now opt to continue these pregnancies given new treatments. This highlights the importance of registries and their ability to capture live data.

Advances in technologies, as demonstrated by the 100,000 Genomes Project [8] and DDD study [5], have shown that cases of diseases we attributed to ‘common’ diseases, (e.g. bronchiectasis), which were excluded from our study are now recognised as RDs [8]. Of note, the recent publication from the UK 100,000 genome project [8] uses similar disease categories to ours (e.g., cerebral palsy) endorsing their inclusion in our figures.

RD patients used 51.9% of paediatric bed days over the 18 years, and 59.9% of teenage bed days, despite representing only 4.2% of the complete cohort (Table 2). Comparing the number of hospital discharges and length of stay by Orphanet category (Table 3) we see that rare neoplastic diseases had the highest mean number of discharges per patient (30) with a mean length of stay of 2.1 days. This is a reflection of the nature of the illnesses and treatment. In contrast, our largest group of patients, those with rare development defects during embryogenesis, had 4.7 mean discharges per patient, with longer mean length of stay of 4.9 days mainly in the early years.

Table 2.

Numbers of people, discharges and LOS for year 2000 livebirth cohort between 1st January 2000 and 31st December 2017.

	Number of discharges	Number of patients	Mean discharges per patient	Total LOS (days)	Mean LOS (days) per discharge [Per patient]	Median LOS (days)	%male Discharges
Cohort 2000	35,225	12,958	2.7	100,879	2.9 [7.8]	1.0	57.7%
Rare disease	14,454	2164	6.7	52,335	3.6 [24.2]	1.0	56%
Subset	9325	1289	7.2	37,606	4.0 [29.2]	1.0	55.4%

Table 3.

Number of discharges by RD patients, number of patients and LOS analysis for this year 2000 livebirth RD-patient cohort by Orphanet classification in decreasing order of frequency.

Orphanet classification	Number of discharges	Number of people	Mean Total LOS	Discharges per patient	LOS (days) discharge	Median LOS (days)
Rare development defect during embryogenesis	4374	927	21,300	4.7	4.9	1.0
Rare neoplastic disease	3633	121	7546	30.0	2.1	1.0
Rare haematologic disease	1605	104	2850	15.4	1.8	1.0
Rare neurologic disease	1317	322	6136	4.1	4.7	1.0
Rare respiratory disease	616	43	3320	14.3	5.4	2.0
Rare endocrine disease	591	135	1539	4.4	2.6	1.0
Rare systematic or rheumatologic disease	461	101	1336	4.6	2.9	1.0
Rare skin disease	437	97	695	4.5	1.6	1.0
Rare inborn errors of metabolism	360	42	2467	8.6	6.9	2.0
Rare bone disease	227	66	631	3.4	2.8	1.0
Rare eye disease	157	37	390	4.2	2.5	1.0
Rare otorhinolaryngologic disease	138	47	865	2.9	6.3	1.0
Rare cardiac disease	138	55	730	2.5	5.3	1.5
Rare renal disease	93	12	649	7.8	7.0	2.0
Rare infectious disease	93	13	456	7.2	4.9	1.0
Rare gastroenterologic disease	91	22	1055	4.1	11.6	3.0
Rare urogenital disease	25	5	39	5.0	1.6	1.0
Rare hepatic disease	14	6	56	2.3	4.0	3.0
Rare immune disease	a	a	a	a	a	a
Rare intoxication	a	a	a	a	a	a
Rare odontologic disease	a	a	a	a	a	a
Rare circulatory disease	a	a	a	a	a	a
All discharges	14,454	2164	52,335	6.7	3.6	1.0

^aThe case numbers are less than 5 and have been excluded to avoid disclosure issues.

For international comparison, we calculated the prevalence of a subset of RDs defined by Walker et al. [9] in addition to the prevalence of all RDs defined by Orphanet (Table 4). Our study population differs from that of Walker et al., Western Australia (WA) [9] and Chiu et al., Hong Kong (HK) [10] in that ours was an exclusively paediatric cohort. Using the WA subset of RDs [9], our data yielded a higher percentage of RD (2.3%) compared to either WA [9] (2.0%) or HK (2018) [10] (1.5%). If we further extrapolate our Irish data using this RD subset to include the theoretical adult onset RD contribution (30.1% [11]), our dataset would hypothetically yield a cumulative lifetime prevalence of 3.1% of RD, which is double that of the HK study [10]. However nearly one third of the codes defined by Walker et al. are from the disease category of rare development defects during embryogenesis, the group for which the ROI has an increased incidence [4]. The WA cohort was an adult population with less than half the ROI’s rate of development defects (19.1 vs. 53.1%) whereas, interestingly, the rate in the HK population is less than half the ROI’s rate in their overall population (21.9 vs. 53.1%) while HK’s paediatric population rate (48.2 vs. 53.1%) is almost as high as that of the ROI. The higher percentage of RDs in this study could also be attributed to (1) inclusion of rare infectious diseases, (2) differences in subject identification (cross-sectional vs. retrospective cohort), and (3) greater ascertainment in our study from Genetics (12%) and Mortality statistics (6%), which capture non-inpatient RDs. Both Walker et al. [9] and Chiu et al. [10] used previously assigned codes in electronic medical records, which could not be replicated in our study.

Table 4.

Number of discharges for RD patients in a subset of RDs described by Walker et al. [9], number of patients and LOS analysis for this year 2000 livebirth subset of RD-patient cohort by Orphanet classification in decreasing order of frequency.

Orphanet classification	Number of discharges	Number of people	Mean Total LOS	Discharges per patient	LOS (days) discharge	Median LOS (days)
Rare development defect during embryogenesis	3554	678	17,640	5.2	5.0	1.0
Rare neoplastic disease	1682	35	3460	48.0	2.1	1.0
Rare haematologic disease	1300	69	2606	48.0	1.8	1.0
Rare neurologic disease	622	162	3065	3.8	5.0	1.0
Rare respiratory disease	507	38	2945	13.3	5.8	2.0
Rare systematic or rheumatologic disease	350	68	892	5.1	2.5	1.0
Rare endocrine disease	334	57	1016	5.8	3.0	1.0
Rare inborn errors of metabolism	299	27	2098	11.0	7.0	2.0
Rare otorhinolaryngologic disease	135	45	855	3.0	6.3	1.0
Rare bone disease	135	32	407	4.2	3.1	1.0
Rare eye disease	126	24	340	5.2	2.7	1.0
Rare gastroenterologic disease	73	17	981	4.2	13.4	9.0
Rare cardiac disease	52	11	463	4.7	8.9	4.0
Rare hepatic disease	11	5	47	2.2	4.3
Rare immune disease	93	12	649	7.8	7.0	2.0
Rare renal disease	a	a	a	a	a	a
Rare skin disease	a	a	a	a	a	a
Rare urogenital disease	a	a	a	a	a	a
All discharges	9325	1276	37,606	7.3	4.0	1.0

^aThe case numbers are less than 5 and have been excluded to avoid disclosure issues.

As a cohort study, we cannot comment on the validity of the estimate of 300,000 people with a RD living in Ireland [12]. However, knowing that 69.9% of RDs manifest in the paediatric period [11], we can estimate that, using the full Orphacode dataset, a further 978 of this cohort would go on to develop a RD throughout adulthood, giving a cumulative lifetime risk of 5.95% (1 in 17) of developing a RD. This is consistent with international estimates [11].

Our study demonstrates the disparity between the incidence of RDs (4.2%; n = 2283 cases) and their hospital usage (51.9%; n = 52,335 days). Notably, 11.8% (n = 270) of the cohort were deceased by 18 years. A further n = 2013 (3.7%) potentially require or required transition to adult services highlighting the importance of RD transition planning for a substantial number of young adults annually.

This study required manual data capture and assignment of RD codes across disjointed data sources. Service planning requires integrated data systems and live data. As a minimum requirement (1) integrated electronic health care records, (2) implementation of unique patient identifiers, (3) use of ORPHAcodes and, (4) RD registries would allow health care systems to count and plan for RD patients.

Author contributions

Conception and design of the study: SAL, DML, EPT, and EG. Acquisition of data: EG. Drafting the manuscript: EG, SAL, and DML. Revising the manuscript for important intellectual content: DML, SAL, EG, AJW, DNM, and EPT. Approval of the version of the manuscript to be published: EG, EPT, AJW, DNM, DML, and SAL.

Funding

The grant support for this study was provided by Temple Street Foundation, Dublin, Ireland. The funder had no role in the study design, execution, analysis or manuscript preparation. The research was undertaken at Temple Street Children’s University Hospital, Dublin.

Data availability

All grouped data generated or analysed during this study are included in this published article. No individual data are available to protect the recognition of individual patients.

Competing interests

The authors declare no competing interests.

Ethical approval

Permission for this project was obtained from CHI Temple Street ethics committee (ref 2017 RD006) and data were handled in compliance with data protection legislation.

Supplementary information

The online version contains supplementary material available at 10.1038/s41431-022-01144-4.