Several epidemiological studies have been performed to establish the prevalence and incidence of the corneal disease keratoconus (KC). Prevalences ranging from nine to 600 per 100 000 have been reported (Hofstetter 1959; Tanabe et al. 1985). A comprehensive American study from 1986 is often cited in the literature (Kennedy et al. 1986). It registered 86 patients with the diagnosis of KC in Olmsted County (Minnesota) from 1935 to 1982 based on ophthalmoscopy, retinoscopy and keratometry. An incidence of 2.0 per 100 000 citizens per year was calculated. Fifty-three patients in the county had the diagnosis of KC on 31 December 1982, and a prevalence of 54.5 per 100 000 residents was calculated.
The incidence and prevalence of KC in the Nordic countries, including Denmark, have never been estimated. The diagnosis of the disease is registered according to the International Classification of Diseases (ICD10), which is based on a hierarchy of disease categories and subcategories. All hospital diagnoses are reported continuously to the Danish health authorities. This reporting system, in combination with the unique identification system of all individuals, serves as a platform for studying the epidemiology of KC in the Danish general population.
Materials and Methods
In the Danish healthcare system, ophthalmological diseases are managed in the following way. Individuals with visual problems consult their optician or (in more severe cases) are referred to a practising ophthalmologist by their physician. Serious cases are subsequently referred to the hospitals, which are public, free of charge and cover the entire population.
The diagnosis of KC is generically established from monitoring visual acuity, pachymetry, slit-lamp biomicroscopy and videokeratography. Standardized methods such as the Keratoconus Prediction Index (KPI), KISA index (based on K, I-S, AST and SRAX values) and the Keratoconus Severity Index (KSI) index [as suggested by Maeda et al. (1994), Rabinowitz & Rasheed (1999) and Avitabile et al. (2004)] have not been adopted in Denmark.
At birth, each citizen is automatically provided with a unique identification number (CPR number) in the Civil Registration System. This number is used for personal identification concerning most aspects of life, such as healthcare, social security, driver's license, school registration, credit cards, etc.
Patients admitted to hospitals are registered by codes for diagnosis and treatment and these codes are reported continuously to the National Patient Registry under the National Board of Health at the Ministry of the Interior and Health. For this study, the date of first registration of IDC-10 code DH186 was extracted from the database for the years 1995–2005. The request was formulated so that records of patients diagnosed with KC in the fellow eye were not included.
Data from the Danish hospitals comprised an 11 year period from 1995 to 2005 (Fig. 1). Between 37 and 116 patients were first-time diagnosed with KC per year with a mean of 71.7 (± 24.6) patients. Within the same time period, the Danish population was 5 326 404 on average (Statistics Denmark, http://www.dst.dk), resulting in a calculated incidence of 1.3 (71.7 / 5 326 404 × 100 000) patients per 100 000 residents per year [95% confidence interval (CI): 1.0–1.7].
The prevalence in its most simple form can be calculated by multiplying the incidence by the life length. The life length is approximately 78 years (Statistics Denmark), but KC is rarely observed in children. Because KC patients do not contribute considerably to the statistics in the first 14 years of life (Jonkers 1950), the time of contribution is 64 years (78 − 14). The prevalence can then be calculated to 86 (64 × 1.3) individuals with KC per 100 000 residents, based on the incidence reported from the hospitals.
The total number of KC patients in Denmark is approximately 4600 (86 / 100 000 × 5 326 404).
The incidence and prevalence of KC in Denmark is unknown. For this reason we sought to estimate these epidemiological numbers.
An incidence of 1.3 patients per 100 000 residents per year and a prevalence of 86 patients per 100 000 residents were calculated. These results are in accordance with the numbers observed by Kennedy et al. (1986) and with an English retrospective study that found 3.3 patients per 100 000 (95% CI: 2.2–4.7), performed over the years 1996–2000 (Georgiou et al. 2004). The absolute prevalence in Denmark can thus be expected to account for 4600 KC patients, corresponding to almost one-tenth percent of the Danish population.
The calculated prevalence and incidence may be underestimated because only patients diagnosed within the hospitals are included in our investigation. These are considered to be the severe cases; milder forms of KC may be diagnosed by practising ophthalmologists without being referred to hospitals. Such patients do not contribute to the calculations: this could explain why our incidence is a little lower than those estimated by Kennedy et al. (1986) and Georgiou et al. (2004).
In order to examine the extent of this problem, the ophthalmologists in Aarhus County were contacted by letter. In the case of potential KC, 77% (10/13) of ophthalmologists would call in the patient in the following year for control, whereas in the case of advanced KC, 92% (12/13) would refer the patient to Aarhus University Hospital (preliminary data). These answers imply that most of the severe cases are actually registered in the National Patient Registry and thus are included in our calculations. However, it is also implied that the patients with mild symptoms, only attended by their practising ophthalmologist, are not included. This group could contribute considerably to the statistics.
Videokeratography was introduced in most ophthalmological departments in the late 1990s. It is possible that the two peaks (2000–2001) represent additional patients diagnosed with the more sensitive technique. It can be argued that previous cases of irregular astigmatism would then be diagnosed as KC, implying that the calculated prevalence and incidence are overestimated. However, we believe that the argument is more relevant in early diagnosing than in the more noticeable cases attended at the hospitals.
Epidemiological studies of rare diseases can be biased towards inheritance if families are large. KC is often reported in the literature as being inherited (Rabinowitz 1998), but the majority of cases still appear sporadic. To our knowledge, no large KC families exist in Denmark.
In conclusion, the number of KC patients is unknown in Denmark. We have estimated the prevalence and incidence of KC in the Danish population at 86 patients per 100 000 residents and 1.3 patients per year, respectively, based on data from an 11 year period. The total number of KC patients in Denmark was calculated to be at least 4600.
This investigation was funded by Øjenforeningen Værn om Synet (Danish Eye Health Society, Copenhagen, Denmark).