The impact of Jade Goody's diagnosis and death on the NHS Cervical Screening Programme.

OBJECTIVES: In August 2008 the British reality TV star Jade Goody made public her diagnosis of cervical cancer. In February 2009 it was announced that she was terminally ill and she died a few weeks later. A surge in cervical screening attendances associated with these events was widely reported. This paper aims to quantify the size of that effect across England, its duration, and whether it affected some groups of women more than others. SETTING: The Cervical Screening Programme in England. METHODS: Routinely collected statistics for the months around Jade Goody's diagnosis and death were compared with those for other periods. RESULTS: About half a million extra cervical screening attendances occurred in England between mid-2008 and mid-2009, the period during which Jade Goody was diagnosed and died; among these were 370 attendances where the test result was suspected neoplasia. At its peak in March 2009, attendance was 70% higher than expected. Increases were seen in both initial and follow-up screening attendances and in colposcopy attendances, and at all ages, though the magnitude was greater for women aged under 50. A substantially greater proportion of the extra attendances of women aged 25-49 on routine recall occurred in women whose attendance was overdue (28% occurred at 60 months or more) and relatively little represented over-screening (8% had been screened within the last 30 months). CONCLUSIONS: The pattern of increased attendance mirrored the pattern of media coverage of Jade Goody's diagnosis and death. It is likely that the increased screening resulted in a number of lives saved.

A fall-off in cervical screening coverage of younger women in developed countries.

OBJECTIVES: To analyse cervical screening coverage data by age over time in a number of developed countries throughout the world, with specific emphasis on trends for younger women and on age differentials between younger and older women. METHODS: Routinely collected cervical screening statistics and survey data were collected on the proportion of women who have undergone cervical screening with cytology in seven countries in the period 1995 to 2005. RESULTS: Data for the 25-29 age group were examined. Coverage fell in most countries, in three by more than 5 percentage points. In two countries while overall coverage rose in the period, the rise was not as steep in the youngest group of women. Data for each available 5-year age group for the different countries shows a similar gradient in most, regardless of the absolute level of coverage. Although the trend is not uniform in every country, it appears that generally the gap between coverage of younger women and coverage of older women increased, sometimes dramatically, between the mid-1990s and the mid-2000s. CONCLUSIONS: There is a general trend in developed countries towards lower coverage in young women (25-29 years old). No common underlying cause has been clearly identified and there is a need for further studies to investigate the possible reasons for this phenomenon.

Benign breast disease and subsequent breast cancer: English record linkage studies.

BACKGROUND: Benign breast disease (BBD) increases the risk of breast cancer, but details of the relationship would benefit from further study in the UK. METHODS: Analysis of linked statistical abstracts of hospital data, including a cohort of 20 976 women with BBD in an Oxford data set and 89 268 such women in an English national data set. RESULTS: Rate ratios (RRs) for breast cancer, comparing BBD and comparison cohorts in these two data sets, were 2.3 (95% CI: 2.2-2.5) and 3.2 (3.0-3.3), respectively. RRs rose with increasing age at BBD diagnosis and remained elevated for at least 20 years after diagnosis. RRs were particularly high for a relatively small number of cancers occurring in the first few months after BBD diagnosis. CONCLUSIONS: Our findings accord well with those in other large studies, mostly done in the USA, in showing a sustained long-term cancer risk after BBD. They also demonstrate that known long-term risks of disease can be reliably identified from linked routine administrative hospital statistics. Most other studies omit cancers in the first few months after BBD. Such cases-presumably either misdiagnosed or miscoded-merit further study to determine whether in fact they include diagnoses of cancer that were initially missed.

A cohort effect in cervical screening coverage?

OBJECTIVES: The objective of this study is to analyse cervical screening coverage data over time. METHODS: Routinely collected cervical screening statistics, in England, on the proportion of women who have undergone cervical screening with cytology during the preceding five years. The participants included all women residents eligible for cervical screening. RESULTS: Overall coverage remained at about 82% or over between 1995 and 2000. Since 2000, however, coverage has drifted slowly down to just over 80% in 2005. Coverage has long been observed to be related to age. In 2005, the coverage rate was 71% in women aged 25-29, 83% in those aged 35-54 and 75% in those aged 55-64. Comparing coverage by age in the three years--1995, 2000 and 2005--shows broad conformity with this pattern in each of the three years, but overlaid is the fact that at ages below 50, the rate has been falling while at ages above 55, the rate has been rising. The fall in screening coverage appears to be largely a cohort effect, with women born in the 1960s and later being increasingly less likely to participate. CONCLUSIONS: No specific reason for this effect is evident. Action could be targeted at women aged 25-34 to address falling coverage.

Active and passive smoking and the risk of breast cancer in women aged 36-45 years: a population based case-control study in the UK.

Active smoking has little or no effect on breast cancer risk but some investigators have suggested that passive smoking and its interaction with active smoking may be associated with an increased risk. In a population based case-control study of breast cancer in women aged 36-45 years at diagnosis, information on active smoking, passive smoking in the home, and other factors, was collected at interview from 639 cases and 640 controls. Women were categorised jointly by their active and passive smoking exposure. Among never smoking controls, women who also reported no passive smoking exposure were significantly more likely to be nulliparous and to be recent users of oral contraceptives. Among those never exposed to passive smoking, there was no significant association between active smoking and breast cancer, relative risk (RR) of 1.12 (95% confidence interval (CI) 0.72-1.73) for past smokers and RR of 1.19 (95% CI 0.72-1.95) for current smokers, nor was there an association with age started, duration or intensity of active smoking. Compared with women who were never active nor passive smokers, there was no significant association between passive smoking in the home and breast cancer risk in never smokers, RR of 0.89 (95% CI 0.64-1.25), in past smokers, RR of 1.09 (95% CI 0.75-1.56), or in current smokers, RR of 0.93 (95% CI 0.67-1.30). There was no trend with increasing duration of passive smoking and there was no heterogeneity among any of the subgroups examined. In this study, there was no evidence of an association between either active smoking or passive smoking in the home and risk of breast cancer.

Oral contraceptive use and cancer. Findings in a large cohort study, 1968-2004.

We examined cancer incidence in relation to oral contraceptive (OC) use in the Oxford Family Planning Association contraceptive study. The study includes 17032 women, recruited at family planning clinics at ages 25-39 years between 1968 and 1974, who were using OCs, a diaphragm, or an intrauterine device. Follow-up data were available until 2004. OC use was not significantly related to nonreproductive cancer. Breast cancer findings (844 cases) likewise were very reassuring (rate ratio (RR) comparing women ever using OCs with those never doing so 1.0, 95% confidence interval (CI) 0.8-1.1). There was a strong positive relationship between cervical cancer incidence (59 cases) and duration of OC use (RR comparing users for 97+ months with nonusers 6.1, 95%CI, 2.5-17.9). Uterine body cancer (77 cases) and ovarian cancer (106 cases) showed strong negative associations with duration of OC use: RRs for 97+ months of use were 0.1 (95%CI, 0.0-0.4) and 0.3 (95%CI, 0.1-0.5) respectively. This apparent protective effect for both cancers persisted more than 20 years after stopping OCs. Combining data for cancers of the cervix, uterine body and ovary, the age adjusted RR for women ever using OCs compared with those never doing so was 0.7 (95%CI, 0.5-0.8). Beneficial effects of OCs on the gynaecological cancers thus outweighed adverse effects.

Body fatness and physical activity at young ages and the risk of breast cancer in premenopausal women.

We examined the relationship between body fatness, sports participation and breast cancer risk in 1560 premenopausal cases and 1548 controls, from three related population-based case-control studies in the UK. Half of the women with breast cancer were aged less than 36 years at diagnosis. Women who perceived themselves as plump at age 10 years had a relative risk of 0.83 (95% confidence interval 0.69-0.99, P = 0.03) as compared with those who perceived themselves as thin. Self-reported obesity compared with leanness at diagnosis was associated with a relative risk of 0.78 (95% confidence interval 0.56-1.06, P = 0.11). Women who reported having been plump at age 10 years and overweight or obese at diagnosis had a relative risk of 0.75 (95% confidence interval 0.56-1.01, P = 0.06) as compared with those who reported being thin at age 10 years and at diagnosis. Findings for three related measures of body fatness suggested that obesity is associated with a reduced risk of premenopausal breast cancer. There was no association between sports participation and breast cancer risk in these premenopausal women. The relative risk for spending an average of more than 1 h per week in sports compared with less from ages 12 to 30 years was 1.00 (95% CI 0.86-1.16, P = 0.98).

Risk factors for adenocarcinoma and squamous cell carcinoma of the cervix in women aged 20-44 years: the UK National Case-Control Study of Cervical Cancer.

We report results on risk factors for invasive squamous cell and adenocarcinomas of the cervix in women aged 20-44 years from the UK National Case-Control Study of Cervical Cancer, including 180 women with adenocarcinoma, 391 women with squamous cell carcinoma and 923 population controls. The risk of both squamous cell and adenocarcinoma was strongly related to the lifetime number of sexual partners, and, independently, to age at first intercourse. The risk of both types of cervical cancer increased with increasing duration of use of oral contraceptives, and this effect was most marked in current and recent users of oral contraceptives. The risk of squamous cell carcinoma was associated with high parity and the risk of both squamous cell and adenocarcinoma increased with early age at first birth. Long duration smoking (20 or more years) was associated with a two-fold increase in the risk of squamous cell carcinoma, but smoking was not associated with the risk of adenocarcinoma. Further studies are needed to confirm the suggestion from this and other studies of differences in risk related to smoking between squamous cell and adenocarcinomas of the cervix.

Mortality study of 18 000 patients treated with omeprazole.

BACKGROUND: The long term safety of potent gastric acid suppressive therapy has yet to be established. METHOD: General practice record review at a median interval of 26 months followed by retrieval of details of all deaths within four years using the UK National Health Service Central Registers in 17 936 patients prescribed omeprazole in 1993-1995. Death rates were compared with general population rates. RESULTS: Records of 17 489 patients (97.5%) were examined. A total of 12 703 patients received further scripts for antisecretory drugs, 8097 for omeprazole only (65.6%): 3097 patients have died. All cause mortality was higher in the first year (observed/expected (O/E) 1.44 (95% confidence intervals (CI) 1.34-1.55); p<0.0001) but had fallen to population expectation by the fourth year. There were significant mortality increases in the first year, falling to or below population expectation by the fourth year, for deaths ascribed to neoplasms (1.82 (95% CI 1.58-2.08); p<0.0001), circulatory diseases (1.27 (95% CI 1.13-1.43); p<0.0001), and respiratory diseases (1.37 (95% CI 1.12-1.64); p<0.001). Increased mortality ascribed to digestive diseases (2.56 (95% CI 1.87-3.43); p<0.0001) persisted, although reduced. Increased mortality rates for cancers of the stomach (4.06 (95% CI 2.60-6.04); p<0.0001), colon and rectum (1.40 (95% CI 0.84-2.18); p=0.075), and trachea, bronchus, and lung (1.64 (95% CI 1.19-2.19); p<0.01) seen in the first year had disappeared by the fourth year but that for cancer of the oesophagus had not (O/E 7.35 (95% CI 5.20-10.09) (p<0.0001) in year 1; 2.88 (95% CI 1.62-4.79) (p<0.001) in year 4). Forty of 78 patients dying of oesophageal cancer had the disease present at registration. Twenty seven of those remaining cases had clinical evidence of Barrett's disease, stricture, ulcer, or oesophagitis at registration (O/E 3.30 (95% CI 2.17-4.80)). Six deaths occurred in patients with hiatal hernia or reflux only (O/E 1.02 (95% CI 0.37-2.22)) and five in patients without oesophageal disease (O/E 0.77 (95% CI 0.25-1.80)). No relationships were detected with numbers of omeprazole scripts received. CONCLUSIONS: Increases in mortality associated with treatment are due to pre- existing illness, including pre-existing severe oesophageal disease. There was no evidence of an increased risk of oesophageal adenocarcinoma in those without oesophageal mucosal damage recorded at registration.

Hormonal replacement therapy, prothrombotic mutations and the risk of venous thrombosis.

Hormone replacement therapy (HRT) increases the risk of venous thrombosis. We investigated whether this risk is affected by carriership of hereditary prothrombotic abnormalities. Therefore, we determined the two most common prothrombotic mutations, factor V Leiden and prothrombin 20210A in women who participated in a case-control study on venous thrombosis. Relative risks were expressed as odds ratios (OR) with 95% confidence intervals (CI95). Among 77 women aged 45-64 years with a first venous thrombosis, 51% were receiving HRT at the time of thrombosis, compared with 24% of control women (OR = 3.3, CI95 1.8-5.8). Among the patients, 23% had a prothrombotic defect, versus 7% among the control women (OR = 3.8, CI95 1.7-8.5). Women who had factor V Leiden and used HRT had a 15-fold increased risk (OR = 15.5, CI95 3.1-77), which exceeded the expected joint odds ratio of 6.1 (under an additive model). We conclude that the thrombotic risk of HRT may particularly affect women with prothrombotic mutations. Efforts to avoid HRT in women with increased risk of thrombosis are advisable.

Using case-control studies for prescribing research.

In this review, the principles of design, analysis and interpretation of case-control studies are illustrated by considering the evidence from case-control studies that link hormone replacement therapy to venous thromboembolism.

Clinical features in women with polycystic ovaries: relationships to insulin sensitivity, insulin gene VNTR and birth weight.

OBJECTIVE: Polycystic ovaries are a common ultrasound finding, yet few of these women have many clinical features of polycystic ovary syndrome. Clinical presentation may relate to degree of insulin resistance, common polymorphism at the insulin gene VNTR, and birth weight. We therefore examined the relationship between insulin sensitivity, insulin gene VNTR genotype, birth weight and presence of polycystic ovaries/features of polycystic ovary syndrome in a normal population study. DESIGN AND PATIENTS: In 224 young women recruited as normal volunteers, ovarian morphology was determined by transabdominal ultrasound and features of polycystic ovary syndrome were identified on clinical and biochemical examination. Insulin sensitivity was estimated from fasting glucose and insulin levels using the homeostasis model. Insulin gene VNTR genotypes were determined in women and their parents. MEASUREMENTS AND RESULTS: Thirty-three per cent (74/224) had polycystic ovaries on ultrasound. These women had higher birth weights (P = 0.004), higher insulin sensitivity (P = 0.02) and higher leptin levels for body mass index (P = 0.04) than women with normal ovaries. However among women with polycystic ovaries, increasing severity of clinical phenotype (based on number of features of: menstrual irregularity, acne, hirsutism, serum testosterone > 3 mmol/l and LH > 10 IU/l) was associated with decreasing insulin sensitivity (P < 0.0001) and related to paternally transmitted insulin gene VNTR class III alleles (P = 0.03). CONCLUSION: Women with polycystic ovaries on ultrasound have increased insulin sensitivity and possible leptin resistance, which could predispose to future weight gain. However, in these women the appearance of clinical features of polycystic ovary syndrome is related to insulin resistance and insulin gene VNTR class III alleles.

The community prevalence of chronic pelvic pain in women and associated illness behaviour.

BACKGROUND: Chronic pelvic pain has often been described as a major women's health issue, but no information exists on the extent of the problem in the United Kingdom. AIM: To investigate the community prevalence of chronic pelvic pain and its effect on the lives of consulting and non-consulting women. DESIGN OF STUDY: Postal questionnaire survey. SETTING: Women aged 18 to 49 (n = 3916) randomly selected from the Oxfordshire Health Authority Register. METHOD: The questionnaire response rate (adjusted for non-deliveries) was 74% (2304/3106). Chronic pelvic pain was defined as recurrent or constant pelvic pain of at least six months' duration, unrelated to periods, intercourse, or pregnancy. Case subgroups comprised recent consulters, past consulters, and non-consulters. Women who reported dysmenorrhoea alone formed a comparison group. RESULTS: The three-month prevalence of chronic pelvic pain was 24.0% (95% CI = 22.1% to 25.8%). One-third of women reported pain that started more than five years ago. Recent consulters (32% of cases) were most affected by their symptoms in terms of pain severity, use of health care, physical and mental health scores, sleep quality, and pain-related absence from work. Non-consulters (41% of cases) did not differ from women with dysmenorrhoea in terms of symptom-related impairment. Irrespective of consulting behaviour, a high rate of symptom-related anxiety was found in women with chronic pelvic pain (31%) compared with women with dysmenorrhoea (7%). CONCLUSIONS: This study showed a high community prevalence of chronic pelvic pain in women of reproductive age. Cases varied substantially in the degree to which they were affected by their symptoms. The high symptom-related anxiety in these women emphasises the need for more information about chronic pelvic pain and its possible causes.

Chronic pelvic pain in the community--symptoms, investigations, and diagnoses.

OBJECTIVES: This study was undertaken to investigate the overlap between chronic pelvic pain, dysmenorrhea, dyspareunia, irritable bowel syndrome, and genitourinary symptoms in the community and also to examine associated investigations and diagnoses. STUDY DESIGN: A postal questionnaire was used to survey 3916 women aged 18 through 49 randomly selected from the Oxfordshire Health Authority Register. The number of responders was 2304 (74% of 3106 questionnaire recipients). Chronic pelvic pain was described as recurrent or constant pelvic pain of > or =6 months' duration unrelated to periods, intercourse, or pregnancy. Case patients (n = 483) were subgrouped as follows: (1) chronic pelvic pain only, (2) chronic pelvic pain and irritable bowel syndrome, (3) chronic pelvic pain and genitourinary symptoms, and (4) chronic pelvic pain, genitourinary symptoms, and irritable bowel syndrome. RESULTS: Half the women with chronic pelvic pain also had either genitourinary symptoms or irritable bowel syndrome, or both. Prevalences of dysmenorrhea and dyspareunia were higher among women with chronic pelvic pain (81% and 41%, respectively) than among women without chronic pelvic pain (58% and 14%, respectively); rates did not differ among the chronic pelvic pain subgroups. Irritable bowel syndrome and stress were the most common diagnoses received by patients with chronic pelvic pain, but 50% had never received a diagnosis. CONCLUSIONS: There is substantial overlap between chronic pelvic pain and other abdominal symptoms in the community. Despite a high prevalence of chronic pelvic pain, many women have never had the condition diagnosed.

Oral contraception and ear disease: findings in a large cohort study.

A number of authors have suggested that oral contraceptives may increase the risk of certain ear diseases, especially otosclerosis and vestibular disorders, although the amount of published information on this topic is limited. We have analyzed the available data on ear disease in the Oxford-Family Planning Association contraceptive study that includes 17,032 women followed for periods of up to 26 years. No evidence of any adverse effect of oral contraceptives on ear disease was detected. A protective effect of oral contraceptives against wax in the ear has been described in the Royal College of General Practitioners oral contraception study. The amount of data available in the Oxford-Family Planning Association study was too small to permit confirmation or refutation of this finding.

Return of fertility in nulliparous women after discontinuation of the intrauterine device: comparison with women discontinuing other methods of contraception.

OBJECTIVE: To clarify the effect of the using the intrauterine device on fertility in nulliparous women. DESIGN: Prospective cohort study of two groups of nulliparous women, one recruited while using an intrauterine device and the other while using an oral contraceptive. SETTING: Seventeen family planning clinics in England and Scotland. SAMPLE: 1,071 nulliparous, married women, aged 18-40 years, 558 of whom contributed information to the main objective of the study. METHODS: The women were recruited between 1982 and 1985 and followed up annually to 1994. Dates and reasons for any contraceptive method changes (which were most frequently to barrier methods) were recorded, together with the outcome of any pregnancies, at each follow up. MAIN OUTCOME MEASURES: The number of nulliparous women giving birth at term after stopping contraception (oral contraceptive, intrauterine device or barrier method) in order to conceive. RESULTS: Women who stopped using a barrier method to achieve a planned pregnancy conceived most quickly: 54% were delivered after one year vs 39% of intrauterine device and 32% of oral contraceptive users (log rank P = 0.002). There was no association between fertility and duration of oral contraceptive use: However, short term intrauterine device users (< 42 months) showed a fertility pattern more favourable than seen in those discontinuing oral contraceptives, with increasing duration of intrauterine device use being associated with decreasing fertility (linear trend P = 0.005); the fertility of women who had used the intrauterine device for 78 + months was the most impaired (28% were delivered by 12 months vs 46% of short term users; at 36 months the corresponding figures were 79% vs 91%). This association remained after adjusting for potential confounding factors, including maternal age, husband's social class, and history of gynaecological illnesses, factors which themselves had independent associations with fertility. CONCLUSIONS: Long term intrauterine device use in nulliparous women appears to be associated with an increased risk of fertility impairment.

Oral contraceptive failures and body weight: findings in a large cohort study.

This report looks at data from 17 032 women who took part in the Oxford Family Planning Association contraceptive study. Thirty-eight first accidental pregnancies occurred during 6779 woman-years of use of progestogen-only oral contraceptives (0.56 per 100) and 95 first accidental pregnancies occurred during 48 692 woman-years of combined oral contraceptive use (0.20 per 100). Although associations with age and parity were found, there was no evidence of any influence of body weight on the risk of accidental pregnancy with either form of oral contraceptive.

The endometrial response to sequential and continuous combined oestrogen-progestogen replacement therapy.

OBJECTIVES: 1. To determine the prevalence of endometrial hyperplasia in postmenopausal women taking standard proprietary regimens of sequential oestrogen/progestogen; 2. to determine the effects of nine months treatment with an oral continuous combined regimen of 2 mg 17beta-oestradiol and 1 mg norethisterone acetate (Kliofem [Kliogest outside the UK]; Novo Nordisk, Denmark) on endometrial histology in postmenopausal women. DESIGN: An open, prospective study in postmenopausal women. SETTING: Fifty-four menopause clinics in the UK. PARTICIPANTS: 2028 postmenopausal women: 1312 (Group A) taking sequential oestrogen-progestogen hormone replacement therapy (HRT), and 716 (Group B) not taking HRT, were recruited. In Group A, 388 women took preparations containing 10 days of progestogen, 921 had 12 days, and 3 had 13 days per cycle. METHODS: Endometrial aspiration biopsies were taken towards the end of a three-month run-in period (Group A) or at study entry (Group B), before administration of the continuous combined HRT regimen. Biopsies were repeated at the end of the nine month treatment period. MAIN OUTCOME MEASURE: Endometrial histology. RESULTS: Initial endometrial biopsy data were available for 1106 women in Group A, who by the time of endometrial investigation had been taking HRT for a median duration of 2.56 years (5th to 95th centiles: 0.77 to 8.49 years). Data were available for 661 untreated women, who had no bleeding and had not taken HRT within the last year (Group B). Complex hyperplasia was found in 59 women (5.3%), and atypical hyperplasia in a further eight (0.7%) in Group A. In Group B there were no cases with complex hyperplasia, but one woman showed atypical hyperplasia (0.2%). At the end of the nine months of continuous combined therapy there was no case of hyperplasia among 1196 biopsies (upper 95% confidence limit of risk 0.31%) in women completing the study. Within this Group all of the women with complex hyperplasia arising during previous sequential HRT and who completed the study (n = 38) reverted to normal endometrial patterns. There was no case of endometrial carcinoma during the study. CONCLUSIONS: Despite taking standard regimens of sequential HRT containing at least 10 days of progestogen, there was a 5.3% prevalence of complex endometrial hyperplasia, and a 0.7% prevalence of atypical hyperplasia. However, continuous combined HRT (Kliofem) containing daily progestogen is not associated with an increased risk of hyperplasia and will convert the endometrium to normal in those with complex hyperplasia arising during previous sequential HRT.

Skin disorders in relation to oral contraception and other factors, including age, social class, smoking and body mass index. Findings in a large cohort study.

BACKGROUND: Over the years, a substantial number of scientific papers has been published considering the possible relationship between oral contraceptive use and various skin disorders. OBJECTIVES: We aimed to investigate the possible effects of oral contraceptives and other factors including age, social class, smoking and body mass index on the occurrence of hospital referral for skin disorders within the Oxford Family Planning Association (Oxford-FPA) contraceptive study. METHODS: We carried out a cohort study with long-term follow up of 17,032 women aged 25-39 years recruited at 17 family planning centres in England and Scotland between 1968 and 1974. The main outcome measure was the pattern of referral to hospital for skin disorders among the participants. The conditions considered (number of women affected in parentheses) were: malignant melanoma (48); other skin cancers (83); benign skin neoplasms (384); boils and cellulitis (334); eczema and dermatitis (327); erythematous disease (54); psoriasis (92); hypertrophic and atrophic conditions (59); diseases of the nails (80); diseases of the hair and hair follicles (74); diseases of the sebaceous glands (362); pigmented naevi (383); and unspecified eruptions (199). Some of these disease categories are broad, which complicates interpretation of the findings. RESULTS: As the data relate to hospital referrals, while most skin disorders are largely managed within primary care, it is difficult to know to what extent our findings can be generalized to disease patterns in the community. None the less, a number of interesting results emerged. These included increased risks (approximately twofold) of referral to hospital for boils and cellulitis in heavy women, for psoriasis in women who smoke, and for pigmented naevi in women of higher social class. Referral for pigmented naevi was reduced by about half in women with a high body mass index. Of the oral contraceptive associations studied, we consider that only an increased risk of referral for eczema or dermatitis in current or recent pill users (relative risk 1.6, 95% confidence interval 1.2-2.1) might reflect a causal relationship. There was no increased risk in past users nor was the risk related to duration of use. These findings are very similar to those reported from the Royal College of General Practitioners oral contraception study. CONCLUSIONS: Hospital referral for certain skin disorders was related to age, social class, smoking and body mass index within the Oxford-FPA study. Hospital referral for skin disorders generally seems to be unrelated to oral contraceptive use within the study, the possible exception being eczema and dermatitis, which occur more commonly in current and recent pill users than in non-users.