Systemic lupus erythematosus prevalence in the U.K.: methodological issues when using the General Practice Research Database to estimate frequency of chronic relapsing-remitting disease

PURPOSE

The purpose of this study was to calculate the prevalence of systemic lupus erythematosus (SLE) between 1992 and 1998 using the General Practice Research Database (GPRD) METHODS: We identified all individuals who had contributed at least 3 years of data to the GPRD and who had a diagnosis of SLE with supporting evidence of their diagnosis. We calculated the annual age- and sex-specific prevalence of SLE. Additionally, we stratified the prevalence by years of data contributed to the GRPD.

RESULTS

In males the point estimate of the prevalence of SLE increased from 7.5/100,000 (CI(95) 6.3, 8.8) to 10.1/100,000 (CI(95) 7.8, 12.2) but this rise was not statistically significant. However, prevalence appeared to increase significantly amongst females from 42.6/100,000 (CI(95) 39.6, 45.6) in 1992 to 70.8/100,000 (CI(95) 65.1, 76.6) in 1998. This increase was mainly amongst women aged 50-79 and in those contributing more than 5 years of data and could not be explained by increasing incidence of SLE or decreasing mortality during the study period.

CONCLUSIONS

We found an increasing prevalence of SLE that could not be explained by increasing incidence or decreasing mortality. This is almost certainly an artefact caused by the increased likelihood of detecting or confirming cases of chronic relapsing-remitting diseases with increasing time contributed to the GPRD.

Venous thromboembolism in pregnancy and the puerperium: incidence and additional risk factors from a London perinatal database

OBJECTIVE

To determine the incidence of venous thromboembolism in pregnancy and the puerperium and to identify risk factors for pregnancy-related venous thromboembolism.

DESIGN

Cohort study and case-control study.

SETTING

London, UK.

POPULATION

395,335 women with live births or pregnancies of 24 or more weeks of gestation between 1988 and 1997.

METHODS

Data extraction from the St Mary's Maternity Information System database. Random sample of 5% for case-control study.

MAIN OUTCOME MEASURES

Incidence of venous thromboembolism; odds ratios for variables associated with venous thromboembolism.

RESULTS

The incidence of venous thromboembolism was 85/100,000 maternities. There were approximately twice as many postpartum as antepartum events. Blood group A, multiple pregnancy, caesarean section, cardiac disease, delivery at gestational age of < 36 weeks, a body mass index of > or = 25, or more and maternal age of 35 or over were all found to increase incidence of venous thromboembolism.

CONCLUSIONS

Although venous thromboembolism is the leading cause of maternal deaths in the UK, it is still a rare event. Most of these events are deep vein thromboses occurring in the postpartum period. Antenatally multiple birth is an important risk factor. Postnatally women who have had a caesarean section, premature delivery or history of cardiac disease should be assessed carefully for venous thromboembolism.

The effects of age, body mass index, smoking and general health on the risk of venous thromboembolism in users of combined oral contraceptives

OBJECTIVES

To investigate the factors associated with idiopathic venous thromboembolism in combined oral contraceptive users and to estimate the crude and age-specific incidence rates ofidiopathic venous thromboembolism among this population.

METHODS

The UK MediPlus Database and the General Practice Research Database were searched to identify women with evidence of venous thromboembolism while exposed to combined oral contraceptives. Cohort and nested case-control studies were carried out using the same methodology on both databases. We conducted a meta-analysis using the individual data for the cases and controls from the two case-control studies to identify factors associated with idiopathic venous thromboembolism in women using combined oral contraceptives.

RESULTS

The incidence rate of idiopathic venous thromboembolism among oral contraceptive users was 39.4 per 100,000 exposed woman-years. The age-specific incidence rates were found to rise sharply after the age of 39 years. Factors identified as being significantly associated with idiopathic venous thromboembolism in women using combined oral contraceptives were: body mass index of 25 kg/m2 and over, the association rising dramatically in women with a body mass index of 35 kg/m2 or more; smoking; general ill health; and asthma.

CONCLUSION

We believe that, before prescribing combined oral contraceptives, the venous as well as the arterial factors need to be considered and, in addition, age, obesity and smoking are all relevant when assessing an individual patient's risk.

Effect of 1995 pill scare on rates of venous thromboembolism among women taking combined oral contraceptives: analysis of general practice research database

OBJECTIVE

To compare the incidence of venous thromboembolism among women taking combined oral contraceptives before and after the October 1995 pill scare.

DESIGN

Analysis of General Practice Research Database.

SETTING

United Kingdom, January 1993 to December 1998.

SUBJECTS

Women aged 15-49 taking combined oral contraceptives.

MAIN OUTCOME MEASURES

Incidence of venous thromboembolism.

RESULTS

Use of so called "third generation" combined oral contraceptives fell from 53% during January 1993 to October 1995 to 14% during November 1995 to December 1998. There was no significant change in the incidence of venous thromboembolism between the two periods after age was adjusted for (incidence ratio 1.04, 95% confidence interval 0.78 to 1.39).

CONCLUSIONS

The findings are not compatible with the assertion that third generation oral contraceptives are associated with a twofold increase in risk of venous thromboembolism compared with older progestogens.