Volume 175 Issue 6

Abstract Objectives: To describe reproductive health issues in women with end-stage renal disease (ESRD) treated with haemodialysis. Study design: Cross-sectional survey based on structured interviews. Setting: Nephrology units of two major metropolitan tertiary referral hospitals in Victoria and their satellite dialysis centres between 1 November 1998 to 30 June 1999. Methods:Women aged 20 years or over in haemodialysis programs. Outcome measures: Menstrual status; prevalence of menstrual and climacteric symptoms; use of gynaecological screening; and prevalence of comorbidities that may benefit from hormone replacment therapy. Results: 48 women completed the survey. They were similar to the 485 women undergoing haemodialysis in Victoria in age (mean age, 55.5 years; range, 20-84 years), years on dialysis (mean age, 3.9 years; range, 1 month-17 years) and primary diagnosis. Eleven of the 15 premenopausal women reported menstrual cycles of 22-35 days, 13 reported common premenstrual symptoms, and six reported dysmenorrhoea that interfered with daily activities. Average age at menopause was 47.7 years (95% CI, 45.6-49.9 years), and six of the 31 postmenopausal women underwent menopause before 45 years. Eight had ever been prescribed hormone replacement therapy (oral in all cases). Over half the women (26) had not had a Pap smear in the last two years, and 12 of those aged over 50 (38%) had not had a mammogram in the same period. Conclusion:Despite their risk of early menopause, cardiovascular disease and bone fracture, few women undergoing haemodialysis were offered hormone replacement therapy. Nor were they adequately screened for gynaecological cancers. Women's health issues seem to be neglected among haemodialysis patients. Chronic renal failure is associated with neuroendocrine disturbances, menstrual disorders and sexual dysfunction,1 and it has been generally accepted that most women undergoing long-term haemodialysis are amenorrhoeic.2,3 However, a North American study noted that the development of strategies to address women's health issues among haemodialysis patients is limited by lack of information about their menstrual patterns, menopausal symptoms, sexual function and use of preventive screening.4 The situation is similar in Australia. As nephrologists increasingly assume the role of primary healthcare providers for haemodialysis patients, with the role of general practitioners less well defined, routine health matters may be overlooked. The problem is increasing as more people enter dialysis programs for end-stage renal disease (ESRD),5 and as their survival improves. Issues unique to women undergoing haemodialysis deserve greater attention. Our aim was to describe the menstrual status, prevalence of menstrual and climacteric symptoms, comorbidities that may benefit from hormone replacement therapy and use of gynaecological cancer screening among women undergoing haemodialysis for ESRD in Victoria. Methods The study was a cross-sectional survey based on structured interviews. Ethics approval was obtained from the research and ethics committees of both Monash Medical Centre and St Vincent's Hospital, Melbourne. Participants Women were eligible if they were aged 20 years or over and were undergoing haemodialysis for ESRD at two tertiary care hospitals (Monash Medical Centre and St Vincent's Hospital, Melbourne) or their associated satellite dialysis centres in Victoria during the survey period, 1 November 1998 to 30 June 1999. Participants were identified from haemodialysis schedules provided by each centre. Exclusion criteria included inadequate spoken English, acute illness and inability to give written informed consent. Survey procedure Participants were interviewed by a psychologist (V S W) at the centre where they underwent haemodialysis (43 women), or, for those undergoing home haemodialysis, by telephone (4) or at home (1). The survey asked about demographic details, history of renal disease, menstrual history, menstrual or menopausal symptoms, pregnancies, gynaecological surgery, sexual function, use of Pap smears and mammograms, medications and comorbidities (eg, diabetes mellitus, cardiovascular disease and osteoporosis). Data analysis Frequencies were compared using the χ2 test (Yates corrected where appropriate) or Fisher's exact test, if the expected cell frequency was less than 5. Results Between 1 November 1998 and 30 June 1999, 73 women aged 20 years and over were listed as undergoing haemodialysis at participating centres. Forty-eight (66%) completed the survey, and 19 were excluded (non-English-speaking [11], too ill [5] or died, changed to peritoneal dialysis or underwent kidney transplantation before interview [1 each]), four were unavailable, and two declined interview. In Victoria as a whole, 485 women aged 20 years or over had haemodialysis in the study period.4 Participant characteristics The 48 participants were aged 20 to 84 years. Their age distribution and documented cause of ESRD matched closely those of the female haemodialysis population in Victoria5(Box 1). Average duration of dialysis was 3.9 years (range, 1 month to 17 years), also matching the average duration for female haemodialysis patients in Victoria (mean, 3.9 years; range, 1 month to 25 years).5 Twenty-seven women (56%) were married or in a defacto relationship, 12 (25%) were widowed or divorced, and nine (19%) were single. There are no comparable Victorian data for relationship status. Menstrual or menopausal status Fifteen women (31%) were classified as premenopausal (< 12 months amenorrhoea and no climacteric symptoms), including one who was taking medroxyprogesterone acetate and had irregular menses. Thirty-one women (65%) were classified as postmenopausal (> 12 months amenorrhoea, surgical menopause or irregular cycles, and persistent vasomotor symptoms), while two (4%) were classified as perimenopausal as they had climacteric symptoms but had not been amenorrhoeic for 12 months (ages, 46 and 47 years). Twenty-four women (50%) reported previous gynaecological surgery; 11 (23%) had had a hysterectomy. Premenopausal women: Ten of the 15 premenopausal women reported regular menstrual cycles. For 11 women, length of cycles fell between 22 and 35 days. Duration of bleeding was 3-9 days (12 women), less than 3 days (2, including the woman taking medroxyprogesterone acetate) and longer than 9 days (1). Bleeding was described as moderate or heavy by 14 women. Premenstrual symptoms were reported by 13 of the 15 premenopausal women (Box 2). Seven reported dysmenorrhoea, which interfered with daily activities in six, and for which four took medication. Postmenopausal women: 31 women were postmenopausal, for between 1.5 and 44 years. Two women had had surgically induced menopause (hysterectomy and bilateral oophorectomy). Of the 29 with non-surgical menopause, 27 reported their age at the time. Their average age at menopause was 47.7 years (95% CI, 45.6-49.9 years). Six women experienced early menopause (< 45 years), including one woman with premature menopause at 35 years. Three women experienced menopause after begining haemodialysis (at ages 49 [2 women] and 53 years [1]). Oestrogen-deficiency symptoms are shown in Box 2. Twelve postmenopausal women and one perimenopausal woman reported that their doctors had ever suggested hormone replacement therapy (HRT), and eight (all postmenopausal) had been prescribed this therapy — oral HRT in all cases. Pregnancies Thirty-seven of the 48 participants (77%) reported having been pregnant, and all but two of these had given birth to at least one child. We did not explore the relationship between pregnancy and dialysis. Sexual function Twenty-five women reported being in a sexual relationship (20 of the 24 who were married or in a defacto relationship and five of the 23 who were single, divorced or widowed). Among postmenopausal women, those who had been prescribed HRT were more likely to be in a sexual relationship (7 of 8 versus 9 of 22 not prescribed HRT; P = 0.04). Eight premenopausal women reported being in a sexual relationship; three of these were not using contraception and had not had a tubal ligation or hysterectomy (ages, 34, 41 and 51 years). Of 42 women who responded to the question "Are you experiencing diminished sexual interest?", 14 answered that they were (33%). Three of 15 premenopausal women reported diminished sexual interest, compared with 10 of 25 postmenopausal women, a difference which was not statistically significant (P = 0.30). Four women reported that this was problematic in their relationships (one premenopausal, one perimenopausal and two postmenopausal women). Breast and cervical cancer screening Most recent screening for breast or cervical cancer is shown in Box 3. Twenty-two women (45%) reported that they had had a Pap smear within the previous two years. This was more likely among premenopausal than postmenopausal women (11 of 15 versus 10 of 31; P = 0.02). Three of the 15 women on the current renal transplant list had not had a Pap smear within the previous two years and did not report having a hysterectomy (ages, 20, 38 and 57 years). Ten women (21%) reported never having a Pap smear. Breast self-examination was performed regularly by 24 of the 36 women who had been shown how to do this by a medical practitioner, and by none of the 12 women who had never been shown (χ2= 13.44; P < 0.001). Of the 31 women aged 50 years or over, 19 (62%) had had a mammogram within the previous two years. Five women reported a family history of breast cancer, and three of these had had a mammogram within the previous two years, while one (aged 50) had never had a mammogram, and the other (aged 79) had had a mammogram over 10 years before. Medications and comorbidities All women were taking some form of medication. Forty-six of the 48 were taking caltrate or calcitriol, and 37 were taking erythropoietin. Fourteen women were being treated for hyperlipidaemia, and 26 for hypertension. Seven women were current cigarette smokers. Ten women (21%) had diabetes mellitus, and 18 (38%) had cardiovascular disease. Prevalence of cardiovascular disease was greater in those with a family history (14 of 24 versus 4 of 24 with no family history; χ2= 7.20; P = 0.007), but was not increased among those with diabetes (5 of 10 versus 13 of 38 without diabetes; P = 0.47). Fourteen women had a history of bone fracture, 11 of whom were postmenopausal. Only one postmenopausal woman with past fracture was taking HRT at the time of the fracture, and two had previously taken HRT. None had been prescribed a bisphosphonate. Risk of fracture was higher in women with a family history of osteoporosis (6 of 8 versus 8 of 40 with no family history; P = 0.005). Discussion We found that the average age of menopause among women undergoing haemodialysis in our study was 47.7 years. About a third of the women were premenopausal, and most of these had menstrual cycles which appeared ovulatory on the basis of cycle length and regularity and presence of premenstrual symptoms.6A third of women who answered a question about sexual interest reported that it was diminished. More than half the women had not had a Pap smear in the previous two years, and more than a third of those aged over 50 years had not had a mammogram in this period. Although our sample was small, it was representative of the adult female haemodialysis population of Victoria in terms of age, cause of ESRD and time on dialysis. Our finding that most premenopausal women had apparently ovulatory menstrual cycles accords with results of a 1997 study of North American women.3 In contrast, earlier studies suggested a high rate of amenorrhoea among women with ESRD, and this difference has been attributed partly to the introduction of recombinant human erythropoietin to correct the anaemia of chronic renal failure in the late 1980s.1Erythropoietin reduces prolactin levels7 and may thereby restore ovulation and improve sexual function. Correction of anaemia, and consequent improved well-being, appetite and nutritional status, could also contribute to improved reproductive function.8,9 Greater emphasis over the past decade on delivery of adequate dialysis may have similar effects, although definitive evidence is lacking. Women with ESRD tend to undergo menopause earlier than healthy women, at an average age of 4710 versus 51.5 years.11 Mean age in our participants (47.7 years) is consistent with previous findings.10,4 The hysterectomy rate of 23% in our participants reflects the rate reported for Australian women overall.12 Sexual dysfunction is a feature of chronic renal failure, with many women complaining of decreased libido and inability to achieve orgasm.11 In our study, a third of responding women reported diminished sexual interest. This occurred despite the high rate of erythropoietin use. Clearly, the pathogenesis of sexual dysfunction is complex, involving not only hormonal factors, but also psychological concerns, body image, nutritional status and comorbid medical conditions. It appeared that sexual issues and contraception among these women were not being adequately addressed. Another concern is that a large proportion of women in this group had not been screened for gynaecological cancers according to current guidelines. In Australia, all women are advised to have a Pap smear every two years from the age of 18, or commencement of sexual activity, until the age of 70.13 Screening mammography for breast cancer is recommended two-yearly for women between the ages of 50 and 70 years.14 Although these guidelines are not followed by all women in the general population, women undergoing haemodialysis have increased risk of gynaecological malignancies,10 which may indeed increase further with immunosuppression after transplantation. The failure of three women on the transplant waiting list to have recommended cervical screening suggests poor clinical practice. The leading cause of death among postmenopausal women with ESRD is cardiovascular disease (42% of all deaths).5 Nearly 38% of our study population reported having cardiovascular disease, which was significantly associated with family history, but not diabetes. The extent to which the excessive cardiovascular mortality of ESRD is a consequence of oestrogen deficiency exacerbating the adverse lipoprotein lipid profile of the anephric state requires investigation.15 In addition, a third of postmenopausal women reported a history of fracture and are at considerable risk of recurrent fracture.16 The role of HRT in postmenopausal women undergoing dialysis is uncertain. While benefits would be expected in terms of fracture prevention, they must be weighed against possible complications, especially thrombosis. In particular, patients bearing fistulas made from artificial materials (eg, polytetrafluoroethylene) appear to have a higher risk of thrombosis if their haemoglobin concentration is higher than 12g/dL.17The risk may be further exacerbated by routine erythropoietin use. Women on maintenance haemodialysis have impaired oestrogen clearance,15 and hence oestrogen therapy should be low dose. Transdermal therapy is preferable to oral therapy, as it is less likely to be procoagulant,18 and circulating levels can be monitored.15 All women treated with HRT in our study were taking it orally. None used local vaginal oestrogen, which, considering the high frequency of sexual problems, is surprising. Dialysis patients have complex health problems requiring specialist care which, combined with their frequent hospital attendance, can result in neglect of the routine health management normally undertaken in general practice. Our findings highlight the need for comprehensive well-woman care programs in dialysis units, incorporating cancer screening, sexual counselling, contraceptive advice, menopausal management and fracture prevention. References Lim VS, Henriquez C, Sievertsen G, Frohman LA. Erythropoietin causes hormonal changes in haemodialysis patients? Ann Intern Med 1980; 93: 21-27. Perez RJ, Lipner H, Abdulla N, et al. Menstrual dysfunction of patients undergoing chronic haemodialysis. Obstet Gynecol 1978; 51: 552-555. Lim VS, Henriquez C, Sievertsen G, Frohman LA. Ovarian function in chronic renal failure: evidence suggesting hypothalamic anovulation. Ann Intern Med 1980; 93: 21-27. Holley JL, Schmidt RJ, Bender FH, et al. Gynaecologic and reproductive issues in women on dialysis. Am J Kidney Dis 1997; 29: 685-690. Disney APS, Russ GR, Walker R, et al. ANZ DATA Registry Report 1998. Adelaide: Australian and New Zealand Dialysis and Transplant Registry, 1999. Mortola JF. Premenstrual syndrome. Curr Ther Endocrinol Metab 1997; 6: 251-256. Schaefer RM, Kotot F, Wernze H, et al. Improved sexual function in haemodialysis patients on recombinant erythropoietin: A possible role for prolactin. Clin Nephrol 1989; 31: 1-5. Tarng DC, Huang TP, Doong TI. Improvement of nutritional status in patients receiving maintenance haemodialyis after correction of renal anemia with recombinant human erythropoietin. Nephron 1998; 78: 253-259. Steffenson G, Aunsholt NA. Does erythropoietin cause hormonal changes in haemodialysis patients? Nephrol Dial Transplant 1993; 8: 1215-1218. Gipson D, Katz LA, Stehman-Breen C. Principles of dialysis: special issues in women. Semin Nephrol 1999; 19: 140-147. Palmer B. Sexual dysfunction in uremia. J Am Soc Nephrol 1999; 10: 1381-1388. Dennerstein L, Shelley J, Smith AM, Ryan M. Hysterectomy experience among mid-aged Australian women. Med J Aust 1994; 161: 311-313. National Health and Medical Research Council. Cervical cancer screening: interval of screening. Canberra: Commonwealth of Australia, 1991. Baker P, Raineri T, Nichols A. Increasing breast and cervical screening rates. Curr Ther 1999; 40 (12): 30-33. Ginsburg ES, Owen W, Greenberg L, et al. Estrogen absorption and metabolism in women with endstage renal failure. J Clin Endocrinol Metab 1996; 81: 4414-4417. Cummings SR, Nevitt M, Browner WS, et al. Risk factors for hip fracture in white women. N Engl J Med 1995; 332: 767-773. Watschinger B, Watzinger U, Templ H, et al. Effect of recombinant erythropoietin on anterior pituitary homones on chronic haemodialysis. Horm Res 1991; 36: 26. Scarabin P-Y, Allhene-Gelas M, Plu-Bureau G, et al. Effects of oral and transdermal estrogen/progesterone regimens on blood coagulation and fibrinolysis in postmenopausal women. Arterioscler Thromb Vasc Biol 1997; 17: 3071-3078. (Received 27 Nov 2000, accepted 14 May 2001) Authors' details Jean Hailes Foundation Research Unit, Melbourne, VIC. Christina Jang, MB BS, Research Registrar; Robin J Bell, PhD, FAFPHM, Consultant Epidemiologist; Vikki S White, Grad Dip Appl Psych, MPH, Research Coordinator; Susan R Davis, PhD, FRACP, Director of Research, and Associate Professor, Department of Preventive Medicine, Monash University, Melbourne, VIC. St Vincent's Hospital, Melbourne, VIC. Petrova S Lee, FRACP, Nephrologist; Karen M Dwyer, MB BS, Nephrology Registrar. Monash Medical Centre, Melbourne, VIC. Peter G Kerr, PhD, FRACP, Deputy Director of Nephrology. Reprints will not be available from the authors. Correspondence: Dr S R Davis, The Jean Hailes Foundation Research Unit, 173 Carinish Road, Clayton, VIC 3168. suedavisATnetlink.com.au Make a comment

Christina Jang · Robin J Bell · Vikki S White · Petrova S Lee · Karen M Dwyer · Peter G Kerr · Susan R Davis

MJA 2001; 175: 320-323 Abstract - Prevalence of smoking in pregnancy - Healthcare provider performance - Efficacy of interventions - Quit-smoking interventions for pregnant women - Nicotine replacement therapy - Bupropion - Conclusion - Acknowledgement - References - Authors' details - - - More articles on Public and environmental health Abstract Smoking doubles the risk of having a low-birthweight baby and significantly increases the rate of perinatal mortality and several other adverse pregnancy outcomes. The mean reduction in birthweight for babies of smoking mothers is 200 g. High quality interventions to help pregnant women quit smoking produce an absolute difference of 8.1% in validated late-pregnancy quit rates. If abstinence is not achievable, it is likely that a 50% reduction in smoking would be the minimum necessary to benefit the health of mother and baby. Healthcare providers perform poorly in antenatal interventions to stop women smoking. Midwives deliver interventions at a higher rate than doctors. The efficacy of nicotine replacement therapy has not been established in pregnancy. Currently, its use should only be considered in women smoking more than 10 cigarettes per day who have made a recent, unsuccessful attempt to quit and who are motivated to quit. Relapse prevention programs have shown little success in the postpartum period. The relationship between smoking and birthweight has been heavily investigated, with studies involving over 500 000 births published by 1986.1 By 1990, the evidence was strong enough for the US Surgeon General to conclude that maternal smoking definitely retarded fetal growth, causing an average reduction in birthweight of 200 g and doubling the risk of having a low-birthweight baby.2 Women who quit smoking in the first 3-4 months of pregnancy have infants of equivalent birthweight to those who have never smoked.2 The relationships between smoking and other adverse pregnancy outcomes have also been widely studied. An Australian group synthesised studies on 10 pregnancy-related diseases and found that smokers had significantly increased risk of eight negative outcomes, including almost treble the rate of sudden infant death syndrome among their offspring (Box 1).3 A more recent study found that smoking was strongly associated with an elevated risk of placenta praevia, abruptio placentae, ectopic pregnancy and preterm premature rupture of the membranes (Box 1).4 Randomised trials of the effects of quitting indicate that abstinence is associated with a reduction in low birthweight (pooled OR, 0.80; 95% CI, 0.67-0.95) and in the incidence of preterm birth (pooled OR, 0.83; 95% CI, 0.69-0.99).5 Given the very low power of relevant trials, it is not surprising that differences in other, less common negative pregnancy outcomes were not detected. If maternal smoking in pregnancy has a lasting impact on the child's long-term physical and mental development, the effect is likely to be small and of uncertain clinical significance.6 One recent review found that effects on children's physical development were related to maternal cigarette consumption during pregnancy, suggesting a dose-response relationship.7 For intellectual development, Fried emphasised that smoking and other drug use accounted for less than 5% of the variance associated with cognitive outcome, whereas other lifestyle factors accounted for up to 35% of this variability.8 Prevalence of smoking in pregnancy International data do not indicate a clear trend towards reduced smoking in pregnancy in developed countries.9 Sweden, a country which has conducted a sustained campaign to reduce smoking by pregnant women, provides a rare example where a decline has been systematically documented.10 In 1983, 29% of pregnant Swedish women smoked daily, and by 1997 this had reduced to 15%.10 In Australia, there have been no long-term serial data on trends in smoking in pregnancy.9 Recent data from New South Wales indicate a small decrease from 22.1% in 1994 to 19.8% in 1998.11 However, smoking prevalence may be much higher in women attending public hospital antenatal clinics, with three Australian studies published in the 1990s reporting proportions of 35%-38%.9 While up to one in three smokers quit when pregnancy is planned or after confirmation of pregnancy,9 research indicates that many pregnant women who smoke are not especially motivated to quit. Intervention programs should focus on increasing readiness to quit.12 Healthcare provider performance Healthcare providers are well placed to inform and influence pregnant smokers to quit, but many do not routinely deliver quit smoking interventions.13 In Australia, most pregnant smokers are identified, but only a minority are counselled about how to quit.14 Nurses report offering more counselling on quitting smoking than general practitioners and obstetricians.14 In a UK study of midwives, general practitioners and obstetricians, most reported experiencing difficulty and lack of enjoyment when giving counselling on quitting smoking.15 Just 28% of these practitioners thought they possessed the necessary skills and 53% perceived themselves to be insufficiently trained. Australian research also suggests that lack of training is a barrier, together with insufficient staff and lack of institutional smoking intervention policies.14 In the mid-1990s, very few Australian antenatal clinics had relevant written policies (4%) or offered in-service training in quit-smoking interventions (12%).14 Despite the difficulties, there is encouraging evidence that antenatal staff are more likely to provide smoking cessation counselling than healthcare providers in other settings.16 Training and institutional support can build on this platform.17 Efficacy of interventions A Cochrane review of 34 trials reported a significant reduction in smoking in the intervention groups (odds ratio, 0.53; 95% CI, 0.47-0.60),5 an absolute difference of 6.4% in the number of women continuing to smoke. In the eight trials with high-intensity interventions, high-quality study methodology and validated measurement of smoking cessation, the resulting absolute difference in quit rates was 8.1%. The prevention of relapse after pregnancy remains problematic. More than two out of three women who quit during pregnancy resume smoking by six months postpartum.18 Five trials of smoking relapse prevention pooled in the Cochrane review showed no significant difference in favour of treatment.5 Recently, biochemical measurements have been recommended as indicators of harm reduction in pregnant smokers.19 One study found a 92 g increase in mean birthweight for babies of women with a reduction of 50% or more in saliva cotinine levels.19 However, the value of such approaches is likely to be vigorously debated. Hebel et al found that the benefits of decreased smoking for birthweight were almost entirely restricted to those who quit completely.20 On the other hand, a recent large Australian trial of a quit-smoking intervention that found no significant effect on quit rates did find that babies born to women in the intervention group were on average significantly heavier (84 g) than the babies of smoking women in the control group, which suggests that reduced smoking may have a positive effect.21 Regardless of the debate about goals, the evidence in favour of routinely counselling pregnant women to quit smoking is sufficiently strong for the US Preventive Services Task Force to give it the highest category recommendation (Category A).22 This is higher than the recommendation given to other common clinical activities, such as routine screening for iron-deficiency anaemia in pregnancy (Category B) or for gestational diabetes mellitus (Category C).22 Quit-smoking interventions for pregnant women Although effectiveness has been most clearly documented for cognitive behavioural programs,13 it is difficult to make recommendations about what specific components should be included in an intervention because of the diversity of methods tested.5 Group programs are very poorly attended and are not recommended.13 Self-help materials appear to be useful, especially when used in the populations for whom they were developed.13,16Box 2 summarises what is known about the effectiveness of different components. Without being inappropriately prescriptive, four main components of smoking cessation interventions are recommended in pregnant and non-pregnant populations alike: assessment, advice, assistance with quitting and follow-up. Assessment Mullen et al have shown that simple changes in question format (away from requiring yes/no answers towards allowing responses such as "I used to smoke" or "I have cut down") increase smoking disclosure.23 Women who report quitting recently have higher rates of inconsistent biochemical measures of smoking, so more detailed questioning of these women is appropriate. Advice Advice should provide specific risk information tailored to the woman's current state of knowledge. For example, women who reject the scientific evidence, countering it with anecdotal information, will require fuller discussion to overcome their misconceptions. Analogies about the effects of nicotine, carbon monoxide and tar may help to clarify the risks. It is in the healthcare provider's interest to record in the notes that advice to quit smoking has been given. Assistance This is probably the most important component, but also the one most commonly omitted by antenatal care providers.14 A useful starting point is to check whether the woman has already attempted to quit and is contemplating attempting to quit now. If the woman is not contemplating quitting, an attempt should be made to identify why and to respond and motivate appropriately. There may be a need to counter specific self-exemptions, such as "It's good to have a smaller baby" or "I only smoke low tar cigarettes", or to address specific concerns, such as the need for smoking as relaxation. In women who are ready to stop, it is valuable to negotiate a target quit date. People who set a definite quit date are most likely to make a serious attempt.24 Subsequent counselling should aim to develop some behavioural aids to quitting (Box 3). Specific concerns about quitting, such as fear of weight gain or withdrawal symptoms, may need to be addressed. Self-help materials can help in this counselling by saving time and reinforcing advice. Many women are situational smokers and will need help to cope with those situations. The woman's partner plays a crucial role in influencing the success of quitting.25 A US study of 688 pregnant smokers found that women whose partners were non-smokers were significantly more likely to have quit smoking prenatally or in early pregnancy than women with smoking partners (58% and 35%, respectively).25 Box 4 summarises an approach to promoting smoking cessation in pregnancy. Follow-up Because of the high frequency of relapse, follow-up, especially at the first postpartum visit, would appear to be a crucial component of a quit-smoking intervention. Unfortunately, in the absence of clear evidence of effectiveness, it is difficult to argue that doctors should invest substantial time in follow-up. Complementary efforts by those providing postnatal services may be required. Nicotine replacement therapy The efficacy of nicotine replacement therapy, combined with at least minimal behavioural assistance, has been well established in many non-pregnant populations.26 Now that nicotine replacement therapy, both gum and patches, is available over-the-counter in Australia, some pregnant women would almost certainly be using it on their own initiative. However, such use remains controversial, as animal models have demonstrated that injected nicotine can cause fetal hypoxia and neonatal death, and has teratogenic effects.27 Animal studies also suggest that there are higher rates of fetal resorption with continuous than with episodic nicotine exposure.27 Clearly, therefore, there is a need to balance the pros and cons of nicotine replacement therapy in pregnancy. Four small "physiological" studies cited by Oncken et al27 did not identify any short-term deleterious effects on the fetus associated with nicotine replacement therapy, except for one unexpected finding where there was a high proportion of loss of fetal heart rate reactivity in the patch condition versus smoking (5/8 versus 1/6). The only published trial of nicotine replacement therapy in pregnancy reported a high overall quit rate (26%) but no differences between intervention and placebo control groups.28 However, the mean birthweight difference was 186 g higher (95% CI, 35-336) in the nicotine replacement therapy group. Nicotine replacement therapy should only be considered in women smoking more than 10 cigarettes per day who have made a recent, unsuccessful attempt to quit and who are motivated to quit. Bupropion The use of antidepressants, including bupropion, for smoking treatment is not recommended in pregnancy. To date, there have been no published studies on their efficacy or risk-benefit ratio in pregnant women. Although the manufacturer is maintaining a database of cases where the most heavily promoted form of bupropion has been used in this population, the numbers are too small (n = 56) to draw any inferences at this stage. Bupropion is classified as a Category B2 drug with respect to use in pregnancy (ie, safety not established). Conclusion The efficacy of cognitive behavioural therapy for quitting smoking in pregnancy has been established. However, healthcare providers perform poorly in this area, particularly in providing counselling about how to stop. Training and institutional support is needed to improve their performance. The effort is well worthwhile: smoking is the most important modifiable cause of adverse pregnancy outcomes in developed nations.2 Acknowledgement The preparation of this manuscript was funded by the New South Wales Cancer Council, Cancer Education Research Program. References Lumley J. Stopping smoking. Br J Obstet Gynaecol 1987; 94: 289-294. US Department of Health and Human Services. The health benefits of smoking cessation. Rockville, MD: Office on Smoking and Health, 1990. (DHHS Publication No. (CDC) 90-8416.) English DR, Holman CDJ, Milne E, et al. The quantification of drug caused morbidity and mortality in Australia. Canberra: Commonwealth Department of Human Services and Health, 1995. Castles A, Adams K, Melvin CL, et al. Effects of smoking during pregnancy. Five meta-analyses. Am J Prev Med 1999; 16: 208-215. Lumley J, Oliver S, Waters E. Interventions for promoting smoking cessation during pregnancy (Cochrane review). In: The Cochrane Library, Issue 2, 2000. Oxford: Update Software, 2000. Walsh RA. The effects of maternal smoking on adverse pregnancy outcomes: an examination of the criteria of causation. Human Biol 1994; 66: 1059-1092. Lassen K, Oei TPS. Effects of maternal cigarette smoking during pregnancy on long-term physical and cognitive parameters of child development. Addict Behav 1998; 23: 635-653. Fried PA. Clinical implications of smoking: determining long-term teratogenicity. In: Zagon IS, Slotkin TA, editors. Maternal substance abuse and the developing nervous system. San Diego: Academic Press, 1992: 77-96. Walsh RA, Redman S, Brinsmead MW, Fryer JL. Predictors of smoking in pregnancy and attitudes and knowledge of risk of pregnant smokers. Drug Alcohol Rev 1997; 16: 41-67. Cnattingius S, Haglund B. Decreasing smoking prevalence during pregnancy in Sweden: the effect on small-for-gestational-age births. Am J Public Health 1997; 87: 410-413. Taylor L, Pym M, Bajuk B, Sutton L, et al. New South Wales mothers and babies 1998. NSW Public Health Bull 2000; Suppl 1: 21. Fava JL, Guise BJ. The trans-theoretical model of smoking: comparison of pregnant and non-pregnant smokers. Addict Behav 2000; 25: 239-251. Walsh R, Redman S. Smoking cessation in pregnancy: do effective programs exist. Health Promot Int 1993; 8: 111-127. Walsh RA, Redman S, Brinsmead MW, Arnold B. Smoking cessation in pregnancy: a survey of the medical and nursing directors of public antenatal clinics in Australia. Aust N Z J Obstet Gynaecol 1995; 35: 144-150. Clasper P, White M. Smoking cessation interventions in pregnancy: practice and view of midwives, GPs and obstetricians. Health Ed J 1995; 54: 150-162. Lowe JB, Balanda KP, Clare G. Evaluation of antenatal smoking cessation programs for pregnant women. Aust N Z J Public Health 1998; 22: 55-59. Cooke M, Mattick RP, Walsh RA. Differential uptake of a smoking cessation programme disseminated to doctors and midwives in antenatal clinics. Addiction 2001; 96: 495-505. McBride CM, Pirie PL. Postpartum smoking relapse. Addict Behav 1990; 15: 165-168. Windsor RA, Qing Li C, Boyd NR, Hartmann KE. The use of significant reduction rates to evaluate health education methods for pregnant smokers: a new harm reduction behavioral indicator? Health Ed Behav 1999; 26: 648-662. Hebel JR, Fox NL, Sexton M. Dose response of birth weight to various measures of maternal smoking duing pregnancy. J Clin Epidemiol 1988; 41: 483-489. Panjari M, Bell R, Bishop S, et al. A randomized controlled trial of a smoking cessation intervention during pregnancy. Aust N Z J Obstet Gynaecol 1999; 39: 312-317. US Preventive Services Task Force. Guide to clinical preventive services, 2nd ed. Baltimore: Williams & Williams, 1996. Mullen PD, Carbonari JP, Takak ER, Glenday MC. Improving disclosure of smoking by pregnant women. Am J Obstet Gynecol 1991; 165: 409-413. Cummings KM, Giovino G, Emont SL, et al. Factors influencing success in counselling patients to stop smoking. Patient Educ Counsel 1986; 8: 189-200. McBride CM, Curry SJ, Grothaus LC, et al. Partner smoking status and pregnant smokers' perceptions of support for and likelihood of smoking cessation. Health Psych 1998; 17: 63-69. Silagy C, Mant D, Fowler G, Lancaster T. Nicotine replacement therapy for smoking cessation. In: The Cochrane Library, Issue 3, 1999. Oxford: Update Software, 1999. Oncken CA, Hardardottir H, Hatsukami DK, et al. Effects of trandermal nicotine or smoking on nicotine concentrations and maternal-fetal hemodynamics. Obstet Gynecol 1997; 90: 569-574. Wisborg K, Henriksen TB, Jespersen LB, Secher NJ. Nicotine patches for pregnant smokers: a randomized controlled study. Obstet Gynecol 2000; 96: 967-971. (Received 18 Dec 2000, accepted 11 May 2001) Authors' details Cancer Education Research Program, NSW Cancer Council, Wallsend, NSW. Raoul A Walsh, BA, DipEd, PhD, Senior Research Academic. Centre for Health Promotion and Cancer Prevention Research, University of Queensland. John B Lowe, DrPH, Director. King Street Medical Practice, Newcastle, NSW. Peter J Hopkins, MB BS(Hons), MMedSc, General Practitioner. Reprints will not be available from the authors. Correspondence: Dr Raoul A Walsh, Cancer Education Research Program, NSW Cancer Council, Locked Bag No 10, Wallsend, NSW, 2287. Make a comment 1: Estimated relative risk of negative outcomes associated with smoking in pregnancy Study Negative outcomes Relative risk (95% CI) Population attributable proportion English et al 19953 Low birthweight* 2.04 (2.03-2.05) 0.23‡ Perinatal mortality* 1.27 (1.21-1.32) 0.07‡ Spontaneous abortion 1.36 (1.32-1.40) 0.09‡ Ectopic pregnancy 1.46 (1.23-1.72) 0.11‡,¶ Antepartum haemorrhage‡ 1.62 (1.56-1.69) 0.15‡ Premature rupture of membranes‡ 1.93 (1.79-2.08) 0.21‡ Sudden infant death syndrome‡ 2.76 (2.66-2.86) 0.34‡ Stillbirth‡ 1.33 (1.27-1.40) 0.09‡ Pre-eclampsia‡ 0.78 (0.76-0.81) -0.07‡ Birth defects 1.01 (0.77-1.33) NA Castles et al 19994 Placenta praevia 1.58 (1.04-2.12) 0.10-0.17§ Abruptio placentae 1.62 (1.46-1.77) 0.11-0.18§ Ectopic pregnancy 1.77 (1.31-2.22) 0.13-0.21§ Preterm premature rupture of membranes 1.70 (1.18-2.25) 0.12-0.20§ Pre-eclampsia 0.51 (0.37-0.63) -(0.09-0.15)§ * Assessment of causality — sufficient evidence. † Assessment of causality — limited evidence. ‡ Based on a prevalence estimate of smoking in pregnancy in Australia of 29% in 1993. ¶ Plus 0.04 due to smoking before conception. § Calculated by one of us (R A W), assuming prevalence of smoking in pregnancy ranged from 20% to 35% in different settings. NA = not appropriate. Back to text 2: Effective and ineffective quit smoking interventions for pregnant women Effective Medical, nursing or other counselling* Self-help materials developed for target population* Ineffective Risk information alone Group behaviour therapy* — very low attendance Self-help materials developed for different population Insufficient evidence Feedback methods Hypnosis — one negative trial Nicotine replacement therapy* — one trial: no effect on quit rates but increased mean birthweight Untested Acupuncture Antidepressants including bupropion* Anxiolytics Aversive smoking therapy Lobeline *Denotes intervention components where reasonable evidence exists for their efficacy in non-pregnant populations Back to text 3: Behavioural tips to quit smoking Tailor these hints to the individual patient. The five Ds Patients may find it useful to recall and practise the five Ds when they feel the urge to smoke: Delay, even for a short while Drink water Deep breathing Do something different and Discuss the craving with another person Other tips Write out a list of reasons to quit and display it prominently (eg, on the fridge) Get rid of all tobacco products, ashtrays, lighters, matches, etc. Clean all clothes to remove cigarette smell. Enlist the support of non-smoking friends, relatives and workmates. Change habits associated with smoking (eg, instead of smoking after meals, chew on a toothpick or change rooms). Change environmental cues. For example, the telephone often causes a ”reflex action” to smoke — so move the telephone to another place to change the cue. Change the daily routine to minimise the association of tobacco with certain activities or times of the day. Keep hands busy (eg, knitting, gardening, drawing, origami). Sit in non-smoking areas. Escape situations that invite a relapse. Talk positively, think positively about quitting and your future. Try to avoid stress immediately after quitting. Substitute another activity for smoking (eg, go for a walk when the urge to smoke strikes). Try daily exercise to keep occupied, to relieve stress, to maintain a positive frame of mind and to become fit. Set aside the money previously spent on cigarettes to buy something as a reward. Do not drink alcoholic beverages, as these are associated with relapse and are best avoided in pregnancy. Avoid, even temporarily, social situations normally associated with smoking. Practise saying, ”No thank you, I don´t smoke”. Ask other smokers not to give cigarettes, offer to buy cigarettes or smoke in your presence. View quitting as a day-at-a-time process rather than an immediate lifelong commitment. Back to text 4: Recommended approach to promoting smoking cessation in pregnancy Back to text

Raoul A Walsh · John B Lowe · Peter J Hopkins