- Email: [email protected]
The Risks of Venous Thromboembolic Disease Among German Women Using Oral Contraceptives: A Database Study
ORIGINAL RESEARCH ARTICLE
The Risks of Venous Thromboembolic Disease Among German Women Using Oral Contraceptives: A Database Study R.D.T. Farmer,* J.-C. Todd,* M.A. Lewis,† K.D. MacRae,* and T.J. Williams* This study investigated the risk of venous thromboembolic disease (VTE) between second and third generation combined oral contraceptives, using the German MediPlus database of patient records. Women studied included 42 patients between the ages of 18 and 49 years, with a diagnosis of VTE treated with an anticoagulant, who were exposed to an oral contraceptive (OC). Four controls per patient (168), matched by year of birth and exposure to an OC on the event day, were identified. More women were users of second generation than third generation OC, and none were using progestogen-only pills. There was no significant difference between patients and control subjects with respect to the type of OC used on the event day (unadjusted odds ratio for third versus second generation users was 0.77; 95% confidence interval [CI] 0.38 –1.57) There was no significant age difference between second and third generation users among patients or control subjects. Between January 1 and the event date, there was no significant difference between the patients and control subjects in terms of the number of oral contraceptive prescriptions, number of consultations for psychotherapeutic complaints, or mixed physical and psychotherapeutic consultations; however, patients did demonstrate significantly more consultations for purely physical complaints compared with control subjects (p ,0.0001). There were no significant consultation differences between patients with pulmonary emboli (n 5 6) and other VTE patients (n 5 36). No significant differences with respect to VTE risk between users of second and third generation oral contraceptives were found in this study. Consultations (physical) for patients were higher than for control subjects before the VTE event. If consultation rate relates to the general health status of a person, this might indicate that VTE risk is *Department of Public Health and Epidemiology, Imperial College School of Medicine, Chelsea and Westminster Hospital, London, UK, and †EPES Epidemiology Pharmacoepidemiology and Systems Research, Berlin, Germany Name and address for correspondence: Richard Farmer, MB, PhD, FFPHM, Professor of Public Health and Epidemiology, Imperial College School of Medicine, Department of Public Health, Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, UK. Tel: 0181 7468160; Fax: 0181 74681 Submitted for publication September 23, 1997 Revised November 7, 1997 Accepted for publication December 5, 1997
© 1998 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
higher among women of poorer health, but that this is not related to the type of progestogen in the oral contraceptive that they use. CONTRACEPTION 1998;57:67–70 © 1998 Elsevier Science Inc. All rights reserved. KEY WORDS:
oral contraceptives, venous thromboembolism, database study, third/second generation, consultation patterns, bias
Introduction
T
hree observational studies published between late 1995 and early 1996 reported odds ratios for venous thromboembolic disease (VTE) among women using so-called third generation oral contraceptives, compared with second generation products, in the range of 1.5–1.8.1–3 Due to the design limitations of all three studies, the interpretation of the results has been questioned.4 – 8 A further study based on the computer records from 147 general practitioners in the UK failed to demonstrate a significant difference in the risk of VTE between oral contraceptives of different generations.9 In order to investigate the issue further, we analyzed the data from a German database.
Materials and Methods This investigation was based on the German MediPlus database, which is similar in principle to both the UK MediPlus database9 and the General Practice Research Database (GPRD).3 It comprises the collated medical records from 451 practices, mainly in the western part of Germany. All of the doctors used a similar computer system and recorded clinical transactions among the patients for whom they were contracted to provide care. Diagnoses are recorded on the database according to the 9th edition of the International Classification of Diseases.10 Potential cases of VTE that occurred between October 1992 and September 1995 among women exposed to an oral contraceptive were identified by searching for all women who had a diagnosis of venous thromboembolic disease (codes: 415.1, 451.0, ISSN 0010-7824/98/$19.00 PII S0010-7824(98)00002-X
68
Farmer et al.
451.1, 451.2, 451.8, 451.9, 453.1, 453.8, and 453.9), a record of at least one prescription for an oral contraceptive, and a record of treatment with an anticoagulant. The complete medical record for each patient who met these criteria was printed and examined individually. The final 42 patients included cases of pulmonary embolism (415.1), deep venous thrombosis (451.1), phlebitis and thrombophlebitis of lower extremities unspecified (451.2), phlebitis/thrombophlebitis of unspecified site (451.9), and venous embolism/thrombosis of unspecified site (453.9). Patients with superficial venous events, those whose anticoagulant therapy was not related in time to the VTE event, those who did not have a prescription for an oral contraceptive covering the event day, and those who had evidence of a previous episode of VTE were excluded. For each patient, four control subjects born in the same year and exposed to an oral contraceptive on the day of the patient’s event were randomly selected from the entire database. For both the patients and control subjects, the oral contraceptive history, recent medical history, the total number of consultations, consultations for contraception, and consultations for psychiatric illness recorded between January 1992 and the event day were abstracted. There were too few women with body mass index (or weight), blood pressure, or smoking habits recorded to warrant the inclusion of these as possible confounding variables.
Results The average number of women who were born between 1943 and 1985 who were registered with the doctors at any point in time was 380,000; however, the number of women on the database increased by 34.7% over the 3 years of the study. There were 106,696 women who had a record of having received one or more oral contraceptive prescriptions between 1992 and 1995. There was a total of 101,797 womenyears of exposure to oral contraceptives during this period. Of 72 potential patients with VTE identified, 30 were excluded because they failed to meet the inclusion criteria, all being nonusers on the event day. Of the 42 cases, 15 were treated with heparin and 27 with warfarin. There were 168 control subjects, who were matched by year of birth and oral contraceptive exposure to the 42 patients. None of the patients or control subjects were using progestogen-only oral contraceptives. There were more women using second generation products than third generation products among both the patients
Contraception 1998;57:67–70
Table 1. Use of oral contraceptives by patients and control subjects Type of Oral Contraceptive Second generation Third generation Others Total
Patients (%)
Control Subjects (%)
27 (64.3) 15 (35.7) 0 (0.0) 42 (100.0)
89 (53.0) 64 (38.1) 15 (8.9) 168 (100.0)
and control subjects (Table 1). All of the products classified as ‘‘others’’ were used by the control subjects and all of these women were using Diane (cyproterone acetate and ethinylestradiol), a product marketed for the treatment of acne. There was no significant difference between the cases and controls with respect to the type of oral contraceptive that was being used on the event day. The unadjusted odds ratio for third compared with second generation users was 0.77 (95% confidence interval 0.38, 1.57). The mean age of the cases using second generation products was 35.4 (standard deviation [SD] 7.69) and 31.1 (SD 7.55) for third generation cases, which was not significantly different (p 5 0.081, unpaired t test). The mean ages of the control subjects were 33.7 (SD 7.99) and 34.5 (SD 7.77) years for second and third generation users, respectively (p 5 0.63). Table 2 shows the mean and median number of oral contraceptive prescriptions and consultations of different types among the patients and control subjects between 1 January 1992 and the event date. There was no significant difference between the patients and control subjects with respect to the number of oral contraceptive prescriptions, number of consultations for psychotherapeutic complaints, and number of consultations for mixed psychotherapeutic and physical complaints. The patients had more than twice the number of consultations for physical complaints preceding the event compared with the control subjects (p , 0.0001). Both patients and control subjects using third generation products had consulted a medical practitioner more often than had those using second generation products, but this difference was not significant. The patients (both second and third generation users) had consulted a medical practitioner on significantly more occasions before the event day than had the control subjects; however, there were no statistically significant differences in any type of consultation between the pulmonary embolism patients (n 5 6) and the other VTE patients (n 5 36).
Contraception 1998;57:67–70
Venous Thromboembolic Disease: German MediPlus Study
69
Table 2. Mean (SD) and median (range) consultations and oral contraceptives prescriptions among patients and control subjects between January 1992 and the event date
Oral contraceptive prescriptions Psychotherapeutic consultations Mixed consultations Physical consultations All noncontraceptive consultations
Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range)
Discussion The use of oral contraceptives in Germany is relatively high. The annual sales are equivalent to 5.8 million women-years, in a population of about 17 million women aged 15 to 45, indicating a point prevalence usage in this age range of about 33%. The corresponding use among the women on the MediPlus database was 31%. Sales figures11 indicate that, during the period of this investigation, 31% of oral contraceptives sold in Germany were third generation products. Among the patients, 36% were third generation oral contraceptive users, compared with 38% among the control subjects. The overall patterns of oral contraceptive use in this investigation are therefore not dissimilar to those of Germany as a whole. The VTE rate among oral contraceptive users was about 4 per 10,000 women-years, which is similar to findings from other population-based studies.12–14 The present study revealed no significant difference between second and third generation oral contraceptives with respect to venous thromboembolic disease. There was a marked difference between patients and control subjects with respect to their consultation patterns before the episode of VTE. The prior consultation rate may be a reasonable proxy for the previous general health of a person. If this is the case, the findings of this investigation may indicate that the risk of VTE is higher among women of poorer general health status, but that this increased risk does not vary according to the type of progestogen in the oral contraceptive that they use. This study, in common with all observational studies, can be criticized. The study population was small and the VTE event under investigation was ill defined; however, the prescription of an anticoagulant provides strong supplementary evidence of the event occurring. In all cases, the medical record was made contemporaneously with the event; therefore it is unlikely that there would be a bias by generation of oral contraceptive. It is noteworthy that the fre-
Cases
Controls
3.07 (1.83) 3 (1–7) 0.29 (0.97) 0 (0–6) 0.26 (0.70) 0 (0–3) 13.57 (11.09) 10 (0–49) 14.12 (11.71) 11 (0–52)
3.35 (2.25) 3 (1–10) 0.27 (2.49) 0 (0–32) 0.11 (0.58) 0 (0–6) 5.44 (7.46) 0 (0–46) 6.8 (8.87) 3 (0–54)
Significance (Mann-Whitney) 0.73 0.27 0.29 ,0.0001 ,0.001
quency of VTE among the users of oral contraceptives found in this study is entirely consistent with that from other studies. The finding that the odds ratio for VTE among users of third generation oral contraceptives against second generation products is not significant is consistent with the findings from the UKbased MediPlus study. In terms of statistical significance, these results are consistent with the results of the Leiden thrombophillia study15 and the UK arms of the World Health Organization (WHO) and Transnational studies.2,4
References 1. World Health Organization. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effect of different progestogens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995;346:1582– 8. 2. Spitzer WO, Lewis MA, Heinemann LAJ, Thorogood M, MacRae KD. Third generation oral contraceptives and risk of venous thromboembolic disorders: an international case-control study. Br Med J 1996;312:83– 8. 3. Jick H, Jick SS, Gurewich V, Myers MW, Vasilakis C. Risk of idiopathic cardiovascular death among non-fatal venous thromboembolism in women using oral contraceptives with differing progestogen components. Lancet 1995;346:1589 –93. 4. Lewis MA, Heinemann LAJ, MacRae KD, Bruppacher R, Spitzer WO. The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Contraception 1996;54:5–13. 5. Farmer RDT. Safety of modern oral contraceptives (letter). Lancet 1996;347:259. 6. Cramer DW. Safety of combined oral contraceptive pills. Lancet 1996;347:546 – 8. 7. Johanisson E. Safety of modern oral contraceptives (letter). Lancet. 1996;347:258. 8. MacRae KD, Kay C. Third generation oral contraceptive pills—is the scare over the increased risk of thrombosis justified? Br Med J 1995;311:1112. 9. Farmer RDT, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Population-based study of risk of
70
10. 11. 12. 13.
Farmer et al.
venous thromboembolism associated with various oral contraceptives. Lancet 1997;349:83– 8. International Classification of Diseases: ICD 9, Volume 1. Geneva: WHO, 1977. Intercontinental Medical Statistics Ltd. IMS International. Vessey M, Mant D, Smith A, Yeates D. Oral contraceptives and venous thromboembolism: findings in a large prospective study. Br Med J 1986;292:526. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, Stadel BV, Lundin FE. Oral contraceptive estrogen dose
Contraception 1998;57:67–70
and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991;133:32–7. 14. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995;15:195–200. 15. Bloemenkamp KWM, Rosendaal FR, Helmerhorst FM, Buller HR, Vandenbroucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a third generation progestogen. Lancet 1995;346: 1593– 6.
The Risks of Venous Thromboembolic Disease Among German Women Using Oral Contraceptives: A Database Study R.D.T. Farmer,* J.-C. Todd,* M.A. Lewis,† K.D. MacRae,* and T.J. Williams* This study investigated the risk of venous thromboembolic disease (VTE) between second and third generation combined oral contraceptives, using the German MediPlus database of patient records. Women studied included 42 patients between the ages of 18 and 49 years, with a diagnosis of VTE treated with an anticoagulant, who were exposed to an oral contraceptive (OC). Four controls per patient (168), matched by year of birth and exposure to an OC on the event day, were identified. More women were users of second generation than third generation OC, and none were using progestogen-only pills. There was no significant difference between patients and control subjects with respect to the type of OC used on the event day (unadjusted odds ratio for third versus second generation users was 0.77; 95% confidence interval [CI] 0.38 –1.57) There was no significant age difference between second and third generation users among patients or control subjects. Between January 1 and the event date, there was no significant difference between the patients and control subjects in terms of the number of oral contraceptive prescriptions, number of consultations for psychotherapeutic complaints, or mixed physical and psychotherapeutic consultations; however, patients did demonstrate significantly more consultations for purely physical complaints compared with control subjects (p ,0.0001). There were no significant consultation differences between patients with pulmonary emboli (n 5 6) and other VTE patients (n 5 36). No significant differences with respect to VTE risk between users of second and third generation oral contraceptives were found in this study. Consultations (physical) for patients were higher than for control subjects before the VTE event. If consultation rate relates to the general health status of a person, this might indicate that VTE risk is *Department of Public Health and Epidemiology, Imperial College School of Medicine, Chelsea and Westminster Hospital, London, UK, and †EPES Epidemiology Pharmacoepidemiology and Systems Research, Berlin, Germany Name and address for correspondence: Richard Farmer, MB, PhD, FFPHM, Professor of Public Health and Epidemiology, Imperial College School of Medicine, Department of Public Health, Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, UK. Tel: 0181 7468160; Fax: 0181 74681 Submitted for publication September 23, 1997 Revised November 7, 1997 Accepted for publication December 5, 1997
© 1998 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
higher among women of poorer health, but that this is not related to the type of progestogen in the oral contraceptive that they use. CONTRACEPTION 1998;57:67–70 © 1998 Elsevier Science Inc. All rights reserved. KEY WORDS:
oral contraceptives, venous thromboembolism, database study, third/second generation, consultation patterns, bias
Introduction
T
hree observational studies published between late 1995 and early 1996 reported odds ratios for venous thromboembolic disease (VTE) among women using so-called third generation oral contraceptives, compared with second generation products, in the range of 1.5–1.8.1–3 Due to the design limitations of all three studies, the interpretation of the results has been questioned.4 – 8 A further study based on the computer records from 147 general practitioners in the UK failed to demonstrate a significant difference in the risk of VTE between oral contraceptives of different generations.9 In order to investigate the issue further, we analyzed the data from a German database.
Materials and Methods This investigation was based on the German MediPlus database, which is similar in principle to both the UK MediPlus database9 and the General Practice Research Database (GPRD).3 It comprises the collated medical records from 451 practices, mainly in the western part of Germany. All of the doctors used a similar computer system and recorded clinical transactions among the patients for whom they were contracted to provide care. Diagnoses are recorded on the database according to the 9th edition of the International Classification of Diseases.10 Potential cases of VTE that occurred between October 1992 and September 1995 among women exposed to an oral contraceptive were identified by searching for all women who had a diagnosis of venous thromboembolic disease (codes: 415.1, 451.0, ISSN 0010-7824/98/$19.00 PII S0010-7824(98)00002-X
68
Farmer et al.
451.1, 451.2, 451.8, 451.9, 453.1, 453.8, and 453.9), a record of at least one prescription for an oral contraceptive, and a record of treatment with an anticoagulant. The complete medical record for each patient who met these criteria was printed and examined individually. The final 42 patients included cases of pulmonary embolism (415.1), deep venous thrombosis (451.1), phlebitis and thrombophlebitis of lower extremities unspecified (451.2), phlebitis/thrombophlebitis of unspecified site (451.9), and venous embolism/thrombosis of unspecified site (453.9). Patients with superficial venous events, those whose anticoagulant therapy was not related in time to the VTE event, those who did not have a prescription for an oral contraceptive covering the event day, and those who had evidence of a previous episode of VTE were excluded. For each patient, four control subjects born in the same year and exposed to an oral contraceptive on the day of the patient’s event were randomly selected from the entire database. For both the patients and control subjects, the oral contraceptive history, recent medical history, the total number of consultations, consultations for contraception, and consultations for psychiatric illness recorded between January 1992 and the event day were abstracted. There were too few women with body mass index (or weight), blood pressure, or smoking habits recorded to warrant the inclusion of these as possible confounding variables.
Results The average number of women who were born between 1943 and 1985 who were registered with the doctors at any point in time was 380,000; however, the number of women on the database increased by 34.7% over the 3 years of the study. There were 106,696 women who had a record of having received one or more oral contraceptive prescriptions between 1992 and 1995. There was a total of 101,797 womenyears of exposure to oral contraceptives during this period. Of 72 potential patients with VTE identified, 30 were excluded because they failed to meet the inclusion criteria, all being nonusers on the event day. Of the 42 cases, 15 were treated with heparin and 27 with warfarin. There were 168 control subjects, who were matched by year of birth and oral contraceptive exposure to the 42 patients. None of the patients or control subjects were using progestogen-only oral contraceptives. There were more women using second generation products than third generation products among both the patients
Contraception 1998;57:67–70
Table 1. Use of oral contraceptives by patients and control subjects Type of Oral Contraceptive Second generation Third generation Others Total
Patients (%)
Control Subjects (%)
27 (64.3) 15 (35.7) 0 (0.0) 42 (100.0)
89 (53.0) 64 (38.1) 15 (8.9) 168 (100.0)
and control subjects (Table 1). All of the products classified as ‘‘others’’ were used by the control subjects and all of these women were using Diane (cyproterone acetate and ethinylestradiol), a product marketed for the treatment of acne. There was no significant difference between the cases and controls with respect to the type of oral contraceptive that was being used on the event day. The unadjusted odds ratio for third compared with second generation users was 0.77 (95% confidence interval 0.38, 1.57). The mean age of the cases using second generation products was 35.4 (standard deviation [SD] 7.69) and 31.1 (SD 7.55) for third generation cases, which was not significantly different (p 5 0.081, unpaired t test). The mean ages of the control subjects were 33.7 (SD 7.99) and 34.5 (SD 7.77) years for second and third generation users, respectively (p 5 0.63). Table 2 shows the mean and median number of oral contraceptive prescriptions and consultations of different types among the patients and control subjects between 1 January 1992 and the event date. There was no significant difference between the patients and control subjects with respect to the number of oral contraceptive prescriptions, number of consultations for psychotherapeutic complaints, and number of consultations for mixed psychotherapeutic and physical complaints. The patients had more than twice the number of consultations for physical complaints preceding the event compared with the control subjects (p , 0.0001). Both patients and control subjects using third generation products had consulted a medical practitioner more often than had those using second generation products, but this difference was not significant. The patients (both second and third generation users) had consulted a medical practitioner on significantly more occasions before the event day than had the control subjects; however, there were no statistically significant differences in any type of consultation between the pulmonary embolism patients (n 5 6) and the other VTE patients (n 5 36).
Contraception 1998;57:67–70
Venous Thromboembolic Disease: German MediPlus Study
69
Table 2. Mean (SD) and median (range) consultations and oral contraceptives prescriptions among patients and control subjects between January 1992 and the event date
Oral contraceptive prescriptions Psychotherapeutic consultations Mixed consultations Physical consultations All noncontraceptive consultations
Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range) Mean (SD) Median (range)
Discussion The use of oral contraceptives in Germany is relatively high. The annual sales are equivalent to 5.8 million women-years, in a population of about 17 million women aged 15 to 45, indicating a point prevalence usage in this age range of about 33%. The corresponding use among the women on the MediPlus database was 31%. Sales figures11 indicate that, during the period of this investigation, 31% of oral contraceptives sold in Germany were third generation products. Among the patients, 36% were third generation oral contraceptive users, compared with 38% among the control subjects. The overall patterns of oral contraceptive use in this investigation are therefore not dissimilar to those of Germany as a whole. The VTE rate among oral contraceptive users was about 4 per 10,000 women-years, which is similar to findings from other population-based studies.12–14 The present study revealed no significant difference between second and third generation oral contraceptives with respect to venous thromboembolic disease. There was a marked difference between patients and control subjects with respect to their consultation patterns before the episode of VTE. The prior consultation rate may be a reasonable proxy for the previous general health of a person. If this is the case, the findings of this investigation may indicate that the risk of VTE is higher among women of poorer general health status, but that this increased risk does not vary according to the type of progestogen in the oral contraceptive that they use. This study, in common with all observational studies, can be criticized. The study population was small and the VTE event under investigation was ill defined; however, the prescription of an anticoagulant provides strong supplementary evidence of the event occurring. In all cases, the medical record was made contemporaneously with the event; therefore it is unlikely that there would be a bias by generation of oral contraceptive. It is noteworthy that the fre-
Cases
Controls
3.07 (1.83) 3 (1–7) 0.29 (0.97) 0 (0–6) 0.26 (0.70) 0 (0–3) 13.57 (11.09) 10 (0–49) 14.12 (11.71) 11 (0–52)
3.35 (2.25) 3 (1–10) 0.27 (2.49) 0 (0–32) 0.11 (0.58) 0 (0–6) 5.44 (7.46) 0 (0–46) 6.8 (8.87) 3 (0–54)
Significance (Mann-Whitney) 0.73 0.27 0.29 ,0.0001 ,0.001
quency of VTE among the users of oral contraceptives found in this study is entirely consistent with that from other studies. The finding that the odds ratio for VTE among users of third generation oral contraceptives against second generation products is not significant is consistent with the findings from the UKbased MediPlus study. In terms of statistical significance, these results are consistent with the results of the Leiden thrombophillia study15 and the UK arms of the World Health Organization (WHO) and Transnational studies.2,4
References 1. World Health Organization. WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effect of different progestogens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995;346:1582– 8. 2. Spitzer WO, Lewis MA, Heinemann LAJ, Thorogood M, MacRae KD. Third generation oral contraceptives and risk of venous thromboembolic disorders: an international case-control study. Br Med J 1996;312:83– 8. 3. Jick H, Jick SS, Gurewich V, Myers MW, Vasilakis C. Risk of idiopathic cardiovascular death among non-fatal venous thromboembolism in women using oral contraceptives with differing progestogen components. Lancet 1995;346:1589 –93. 4. Lewis MA, Heinemann LAJ, MacRae KD, Bruppacher R, Spitzer WO. The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Contraception 1996;54:5–13. 5. Farmer RDT. Safety of modern oral contraceptives (letter). Lancet 1996;347:259. 6. Cramer DW. Safety of combined oral contraceptive pills. Lancet 1996;347:546 – 8. 7. Johanisson E. Safety of modern oral contraceptives (letter). Lancet. 1996;347:258. 8. MacRae KD, Kay C. Third generation oral contraceptive pills—is the scare over the increased risk of thrombosis justified? Br Med J 1995;311:1112. 9. Farmer RDT, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Population-based study of risk of
70
10. 11. 12. 13.
Farmer et al.
venous thromboembolism associated with various oral contraceptives. Lancet 1997;349:83– 8. International Classification of Diseases: ICD 9, Volume 1. Geneva: WHO, 1977. Intercontinental Medical Statistics Ltd. IMS International. Vessey M, Mant D, Smith A, Yeates D. Oral contraceptives and venous thromboembolism: findings in a large prospective study. Br Med J 1986;292:526. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, Stadel BV, Lundin FE. Oral contraceptive estrogen dose
Contraception 1998;57:67–70
and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991;133:32–7. 14. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995;15:195–200. 15. Bloemenkamp KWM, Rosendaal FR, Helmerhorst FM, Buller HR, Vandenbroucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a third generation progestogen. Lancet 1995;346: 1593– 6.