Milstone LM, Miller K, Haberman M, Dickens J. Incidence of Moderate to Severe Ichthyosis in the United States. Arch Dermatol. 2012;148(9):1080-1081. doi:10.1001/archdermatol.2012.1702
Author Affiliations: Department of Dermatology, Yale University School of Medicine, New Haven, Connecticut (Dr Milstone); Thomson Reuters, Analytic Consulting and Research Services, Santa Barbara, California (Mss Miller and Dickens); Thomson Reuters, Pharma/Biotech Custom Analytics, Washington, DC (Ms Haberman). Ms Haberman is now with the Association of American Medical Colleges, Washington, DC.
Most statistics on the incidence of hereditary ichthyosis are acknowledged as best-guess estimates. The exceptions are population-based studies for the prevalence of steroid sulfatase deficiency (x-linked ichthyosis)1 and mutations in filaggrin (ichthyosis vulgaris).2 We wondered whether the unique International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for ichthyosis might be used to gather incidence information from large administrative claims databases. We reasoned that moderate to severe ichthyosis is clinically distinctive and usually diagnosed at birth and that infants would not have manifested ichthyosis vulgaris (the common, mild form of ichthyosis), forms of nonhereditary “acquired” ichthyosis, or other skin diagnoses easily confused with ichthyosis.
Administrative claims and enrollment data from 2001 through 2010 were analyzed from 2 Thomson Reuters MarketScan databases: the Commercial Claims and Encounter Database (CCAE) includes more than 100 million privately insured individuals in employer-sponsored health plans; the Medicaid Multi-State Database (Medicaid) reflects the experience of 30 million enrollees from 13 states. Discharge data from the same period came from the Healthcare Cost and Utilization Project (HCUP) Kids' Inpatient Database (KID) sponsored by the Agency for Healthcare Research and Quality (AHRQ) and accessed through the HCUPnet portal. The records in all databases are de-identified and comply with HIPAA (the Health Insurance Portability and Accountability Act).
For a patient to be included in the analysis, the ICD-9-CM diagnosis of 757.1 must have appeared on at least 1 of their inpatient or outpatient claims from the CCAE or Medicaid databases during the first 6 months of life, and patients with this diagnosis were then compared with all individuals born that year. For the HCUP data, babies with ichthyosis were identified as those discharged with the 757.1 diagnosis during the first year of life and were compared with all discharges during that year. The CCAE and Medicaid databases are corrected to represent unique individuals. The HCUP data include multiple discharges for individuals and cannot be categorized into boys and girls.
During the 10-year period of the study, the average numbers of babies given the diagnosis of ichthyosis in the CCAE and Medicaid databases were 25 and 39 per year, respectively. During this same period, the average respective numbers of individuals younger than 1 year whose records were included in the databases were 288 000 and 404 000 per year. Using HCUP data, we found that the average number of babies discharged with the diagnosis of ichthyosis was 220 of the 4.6 million overall discharges within the age range per year. The incidence of babies with the diagnosis of ichthyosis (Table) averaged 9.2 per 100 000 for the claims databases and 4.8 per 100 000 for the discharge database.
Differences between claims and discharge databases might be explained by the fact that claims databases include patients first given the diagnosis of ichthyosis in any inpatient or outpatient care setting, while HCUP data represent only the inpatient setting. Baby boys had a higher incidence of ichthyosis than girls (except for CCAE data in 2001 and 2003), and the difference appears to be increasing over time. We hypothesize that the sex difference is caused by increased diagnosis of recessive X-linked ichthyosis (RXLI) made by fluorescence in situ hybridization (FISH) following an abnormal triple screen during pregnancy. The incidence of RXLI is thought to range from 15 to 50 per 100 000, based on enzyme and genetic studies.2 The discrepancy between boys and girls in our data does not approach 50 per 100 000, but this may be because FISH is not a routine test in many communities, and RXLI is often a relatively mild form of ichthyosis making it clinically unrecognizable early in life.
What is the best number for incidence of moderate to severe ichthyosis to use from these data? Since girls do not get the milder form of ichthyosis, RXLI, we can take the incidence of girls with the ICD-9-CM 757.1 diagnosis as reasonably representative of moderate to severe ichthyosis. That average incidence over the 10 years of the study is 6.7 per 100 000. An oft-quoted but unreferenced figure of 1 in 200 000 has been used for 2 of the more common and easily recognizable subtypes of disease: lamellar ichthyosis and epidermolytic hyperkeratosis. If we add to those incidence numbers the dozen rarer syndromic and nonsyndromic types of clinically distinctive congenital ichthyosis,3 6.7 in 100 000 for all moderate to severe types of ichthyosis seems quite consistent. To our knowledge, this is the first estimate of moderate to severe ichthyosis based on patient data.
It should be noted that the databases used for this study are de-identified, precluding independent validation of ICD-9-CM coding. Concerns about this limitation are mitigated, however, by the distinctive clinical presentation, the unique code for ichthyosis, and our limiting the sample to the first year of life.
Correspondence: Dr Milstone, Department of Dermatology, Yale University School of Medicine, New Haven, CT 06520 (firstname.lastname@example.org).
Accepted for Publication: April 2, 2012.
Author Contributions: Dr Milstone had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Milstone, Miller, and Haberman. Acquisition of data: Milstone, Miller, Haberman, and Dickens. Analysis and interpretation of data: Milstone, Miller, and Haberman. Drafting of the manuscript: Milstone. Critical revision of the manuscript for important intellectual content: Milstone, Miller, Haberman, and Dickens. Statistical analysis: Milstone. Administrative, technical, and material support: Milstone, Miller, Haberman, and Dickens. Study supervision: Milstone.
Financial Disclosure: None reported.