Jump to: Page Content, Site Navigation, Site Search,
You are seeing this message because your web browser does not support basic web standards. Find out more about why this message is appearing and what you can do to make your experience on this site better.
BMJ 2005;330:126-127 (15 January), doi:10.1136/bmj.38326.606979.79 (published 21 December 2004)
Henrik Toft Sørensen, professor1, Lars Pedersen, biostatistician1, Mette Vinther Skriver, biostatistician1, Mette Nørgaard, senior registrar1, Bente Nørgård, associate professor1, Elizabeth E Hatch, assistant professor2
1 Department of Clinical Epidemiology, Aarhus University Hospital, DK-8000 Aarhus C, Denmark, 2 Department of Epidemiology, School of Public Health, Boston University, Boston, MA, USA
Correspondence to: H T Sørensen hts{at}soci.au.dk
We identified all cases of hypospadias that had a full prescription history in the period 1989-2003 from the Danish hospital discharge registry, which contains all discharges from hospitals in Denmark since 1977 and includes 10 digit personal identifiers, surgical procedures, and up to 20 discharge diagnoses classified according to ICD-8 (international classification of diseases, eighth revision) until the end of 1993 and ICD-10 after 1993. The codes for hypospadias in ICD-8 are 752.20, 752.21, 752.22, 752.28, and 752.29; in ICD-10, the codes are Q54.0, Q54.1, Q54.2, Q54.3, Q54.4, Q54.8, and Q54.9.
We found a total of 319 cases of hypospadias (any time post partum) in the four counties. From the Danish birth registry, which contains information on all births in Denmark since 1 January 1973, we selected a control group of 10 records of male births without a diagnosis of hypospadias and with a full prescription history during the same period. We matched cases and controls for birth month, birth year, and county of residence of the child.
The Danish national health service reimburses part of patients' expenditure on many prescribed drugs, including clomifene. The four counties have pharmacies equipped with electronic systems that record information on the drug, dose, personal identification number, and date of dispensing of the drug. All data are transferred to a research database at Aarhus University Hospital. The data from the four counties are electronically available from 1 January 1989 (North Jutland), 1 January 1996 (Aarhus), and 1 January 1998 (Ringkjoebing and Viborg). We took data on all prescriptions for clomifene during the first trimester and 90 days before conception. To avoid confounding, we also took data about prescriptions for antidiabetic drugs and antiepileptic drugs and we also collected data on pre-eclampsia, as it has been associated with hypospadias. We used conditional logistic regression adjusted for available variables to estimate the relative risk for hypospadias after exposure to clomifene. We found an adjusted odds ratio of 0.48 (95% confidence interval 0.15 to 1.54) for hypospadias associated with clomifene (table). Restricting the exposure to clomifene to the first trimester and up to 30 days before the time of conception did not change the risk estimate substantially.
|
|
Contributors: HTS developed the idea and designed the study together with LP and MVS. LP, MVS, MN, and BN collected and validated the data. LP, MVS, and HTS analysed the data. HTS wrote the first draft. EEH, LP, MVS, MN, and BN critically revised the paper. HTS is guarantor.
Funding: Western Danish Research Forum for Health Sciences.
Competing interests: None declared.
![]()
CiteULike
Complore
Connotea
Del.icio.us
Digg
Reddit
StumbleUpon
Technorati
Twitter What's this?