Does an early postnatal check-up improve maternal health: results from a randomised trial in Australian general practice

Authors

  • Jane Gunn,

    Senior Lecturer , Corresponding author
    1. Department of General Practice and Public Health, University of Melbourne, Australia
      Correspondence: Dr J. Gunn, Department of General Practice and Public Health, University of Melbourne, 200 Berkeley Street, Carlton, Melbourne 3053, Australia.
    Search for more papers by this author
  • Judith Lumley,

    Professor
    1. Centre for the Study of Mothers' and Children's Health, LaTrobe University, Australia
    Search for more papers by this author
  • Patty Chondros,

    Research Assistant
    1. Department of General Practice and Public Health, University of Melbourne, Australia
    Search for more papers by this author
  • Doris Young

    Professor
    1. Department of General Practice and Public Health, University of Melbourne, Australia
    Search for more papers by this author

Correspondence: Dr J. Gunn, Department of General Practice and Public Health, University of Melbourne, 200 Berkeley Street, Carlton, Melbourne 3053, Australia.

Abstract

Objectives To investigate whether a visit to a general practitioner one week after discharge results in less depression, increased breastfeeding rates, improved patient wellbeing, fewer physical problems and greater satisfaction with general practice care than the traditional six week postnatal check-up.

Design A randomised controlled trial.

Setting Rural and metropolitan Victoria, Australia.

Population Women giving birth at one rural and one metropolitan hospital between February and December 1995 inclusive.

Methods All women received a letter and appointment date to see a general practitioner for a check-up: the intervention group for one week after hospital discharge, the control group for six weeks after birth. A mail-out survey was conducted at three and six months after birth, including Edinburgh Postnatal Depression Scale and Short Form 36.

Results 1017/1407 (72.3%) women giving birth at participating hospitals were eligible for the trial: 683 (67.2%) gave informed consent. The average response rate to postal follow up at three and six months was 67.5%. No significant differences were found between the groups in: Edinburgh Postnatal Depression and Short Form 36 scores; number of problems; breastfeeding rates; or satisfaction with general practitioner care. Women in the intervention group were less likely to attend for their check-up (76.4%vs 88.4%; P=0.001), more likely to discuss labour and birth at their check-up (OR=1.77, 95% CI 1.17–2.68), less likely to have a vaginal examination (OR = 0.51; 95% CI 0.34–0.77) or pap smear (OR = 0.34; 95% CI = 0.22–0.52) at their check; more likely to report difficulties with low milk supply (OR= 1.72; 95% CI = 1.12–2.66) and adjusting to the demands of a new baby (OR= 1.76; 95% CI 1.13–2.74), more likely to talk to a general practitioner about their baby (68.2%vs 58.0%; P= 0.02) and less likely to consult a hospital doctor about their baby (7.3%vs 14.0%, P = 0.02).

Conclusions To make clinically important improvements in maternal health more is required than early postnatal review.

INTRODUCTION

Traditionally, women have been advised to have a check-up six weeks after childbirth in the belief that the visit would either prevent or allow the general practitioner (GP) to identify and manage postnatal problems1. A recent UK study showed that more than 90% of women attend a postnatal check-up2. Despite these high levels of uptake, there continue to be concerns about high levels of maternal morbidity after birth. Eighty-five percent of women surveyed by Glazener et al.3 experienced at least one health problem two weeks after childbirth3. It is of particular concern that 12 to 18 months later 76% of women were still experiencing at least one problem. Even when women are troubled by conditions, such as back pain, urinary incontinence, depression, haemorrhoids and perineal pain, they are unlikely to raise these issues with their doctor4.

Interestingly, this failure to raise issues of concern cannot be attributed to a lack of contact with a GP. Women have many encounters with GPs in the early postnatal weeks. An Australian review of visits to medical practitioners showed that the average number of visits for a mother-baby pair was 7.7 visits in the six months after birth, which is higher than the 2.64 visit average for all Australians5. Visits begin in the week after birth and are fairly constant through-out the six months, apart from an increase around the time of the traditional six week check-up10. The high incidence of problems and women's reluctance to raise issues of concern makes postnatal care a particularly challenging area for health professionals.

In Australia postnatal care is provided by maternal and child health nurses, who visit a mother at her home within a few days of discharge from hospital and a medical practitioner (GP or obstetrician) who sees the woman for a postnatal check-up six weeks after the birth. It is not routine practice for a GP to visit a new mother at her home.

A number of authors have called for a review of the timing and/or content of the six week check-up6–10. One of the earliest reasons for recommending a review was concern that a six week check-up was too late to deal with contraceptive issues6. This paper reports on a randomised controlled trial undertaken to address the issue of timing of the routine postnatal check-up. We hypothesised that a change in the timing of the postnatal check-up, to one week after discharge from hospital, would provide GPs with the opportunity to review the mother and baby early in the postnatal phase, and to identify and manage problems early, possibly preventing morbidity. It was left up to each GP to determine whether a six week postnatal check-up was also required. The main objectives were to increase breastfeeding rates, to improve physical and emotional wellbeing and to decrease the number of physical problems in the six months after birth. The early postnatal check-up was a simple, relatively cheap intervention which could be easily implemented within current models of care if shown to be effective.

METHODS

To increase the generalisability of the findings and to facilitate recruitment women were recruited from one metropolitan and one rural hospital in Victoria, Australia. Ethics approval was granted from the University of Melbourne Human Research Ethics Committee and from ethics committees at each of the recruiting hospitals. Written, informed consent was obtained from each participant. To be included in the trial women had to be 1. mother of a liveborn infant of 37 or more weeks gestation not requiring care in a special care or intensive care nursery; 2. a public hospital patient; 3. able to read and write English; 4. willing to visit a GP for a postnatal check-up; 5. able to nominate or choose a GP they would visit; 6. leaving hospital on the same day as their baby; and 7. have no plans to see their GP early. Women excluded from the trial were those who were 1. patients of GPs on the project reference group; 2. attending the teenage clinic for antenatal care (by hospital request); or 3. delivered by emergency caesarean section at the metropolitan hospital only (by request of obstetricians).

The project was undertaken as a randomised controlled trial, stratified by recruiting centre. A variable block randomisation procedure was used to make up the randomisation schedule for each recruiting centre11 Women were approached on the postnatal ward, usually on the second or third day after giving birth. After consenting to the trial the woman was allocated to control or intervention groups via a telephone call to the randomisation centre. The intervention consisted of a referral letter and appointment date for a postnatal visit with a GP of the patient's choice one week after discharge from hospital. The control consisted of a referral letter and appointment date for a postnatal visit with a GP of the patient's choice six weeks after childbirth.

Sample size calculations were made using the breastfeeding rates of 50% at three months and 35% at six months for the Victorian population12. To detect a 15% increase in breastfeeding rates 650 women (allowing for a 70% response rate to mail out surveys) were required (α= 0.05, β= 0.20). This sample size was large enough to detect changes in scores across each domain of the SF 36 which are smaller than those described for GP attenders with mild problems compared with the population normative data in the UK13. The sample size also enabled detection of clinically significant changes in scores on the Edinburgh Postnatal Depression Scale (EPDS) of 0.23 SD, from a mean score of 7.6 (SD 4.8)14 to < 6.5.

Mail-out surveys, with reply paid return envelope, were sent at three and six months postpartum. Follow up of non-responders was by reminder phone call and postcard at two week intervals for each of the surveys. Both surveys contained the EPDS15, SF 3616 and items about physical health and health seeking behaviour used in a Statewide survey of recent mothers (with permission from S. Brown, Centre for Study of Mothers’ and Children's Health, LaTrobe University, Melbourne, Australia). At three months women were asked in detail about their postnatal check-up and at six months about their experience of general practice care over the six months. All women were provided with a card to record any contact they had with health professionals in the six months after birth.

The analysis was undertaken using the intention-to-treat principle, by maintaining the women in their original group allocation whether or not they complied with their visit. Calculations were performed with SAS software (SAS Institute Inc. Release 6.1 for Windows; Cary, North Carolina, USA), using χ2 comparisons and t-tests as appropriate. Survival analysis was used to compare breastfeeding duration.

RESULTS

Recruitment occurred between February and December 1995 and 1407 seven women were approached. Of these, 1017 (72.3%) were eligible and were asked to participate. Six hundred and eighty-three women (67.2%) gave written informed consent to participate in the trial; 300 at the metropolitan centre and 383 at the rural centre. The response rate for the three month follow up survey was 4761683 (69.7%); 66.7% at the metropolitan centre and 72.1% at the rural centre. The response rate for the six month follow up survey was 4461683 (65.3%); 58.7% at the metropolitan centre and 70.5% at the rural centre. There were no significant differences in the response rates for the three or six month surveys between the control and intervention groups. Of the 334 women who declined to participate in the trial, 328 agreed to complete the four-item survey for non-participants. participants and non-participants were not significantly different in age, birthweight, parity or mode of delivery. Seventeen women (2.5%) with-drew from the trial (seven from the metropolitan centre and 10 from the rural centre) after giving informed consent to take part. Six withdrew before the three month follow up, eight withdrew after receiving the three month survey, and three withdrew after responding to the three month survey. Most women stated they withdrew because they did not have time to complete the surveys.

Table 1 shows the similarities in the characteristics of women in the intervention and control groups. Women in the control group were more likely to attend for their postnatal check-up at the scheduled time than women in the intervention group (88.4%vs 76.4%, χ2= 11.9, df = 1; P= 0.001), although most women (96%) did attend a check-up at some time. Women in the intervention group were more likely to discuss their labour and birth experience at the post-natal check-up, but there were no other significant differences (Table 2). Table 3 shows the examination performed.

Table 1.  Demographic characteristics of participants: intervention compared with control group. Values are expressed as n (%) or mean SD.
 Early visit*n= 232Control*n= 243
  1. *No significant differences between the groups on any variable.

Age (years)27.5 [5.3]28.0 [5.2]
Primiparous124(36.5)124(36.3)
Birthweight (g)3439 [467.1]3433 [459.2]
Type of delivery
  Normal253 (74.2)265 (77.5)
  Forceps/assisted35 (10.3)25 (7.3)
  Caesarean section53(15.5)52(15.2)
Breastfeeding at recruitment289 (84.8)285 (83.6)
Have own GP292(85.6)296 (86.6)
Antenatal care with a GP197(57.8)207(60.5)
English as first language304(89.2)300(87.7)
Married/hiving with partner291 (85.4)301(88)
Occupation of woman
  Home duties107(31.7)103 (30.3)
  Professional39(11.6)34 (10)
  Salesperson102(30.3)131 (38.5)
  Factory worker52(15.4)40(11.8)
  Unemployed/pensioner37(11)32(7.4)
Partner with profession58 (19.4)62(20.5)
Table 2.  Discussion at the postnatal check-up for early and control visits. Denominators may vary due to missing values. Values are expressed as n (%).
 Early visit n= 211Control n= 221OR [95% CI]
Labour and birth159(75.4)140 (63.4)1.77 [1.17–2.681
Care in hospital64(31.4)51 (23.5)1.49 [0.97–2.29]
Feeding the baby184(85.2)181(81.9)1.27 [0.77–2.11]
Baby's behaviour/crying/sleeping164(76.6)165(75)1.09 [0.71–1.69]
Contraception180(83.3)189(84)0.95 [0.58–1.57]
Physical problems107(51.2)104(47.5)1.16 [0.79–1.69]
My feelings122(57.6)111(51.2)1.29 [0.89–1.89]
Help with housework45(22)42(19.5)1.16 [0.72–1.85]
Help with baby74(35.8)77(35.5)1.01 [0.68–1.50]
Relationships53 (25.9)51 (23.9)1.11 [0.71–1.72]
Sex63 (30.3)84(38.7)0.69 [0.46–1.03]
Diet47(22.6)51 (23.6)0.94 [0.60–1.48]
Exercise69 (33.2)83 (37.7)0.82 [0.51–1.21]
Childcare23(11.2)22(10.2)1.11 [0.60–2.04]
Going back to work26(12.6)19(8.7)1.51 [0.81–2.79]
Table 3.  Examination at the postnatal check-up for early and control visits. Denominators may vary due to missing values. Values are expressed as n(%).
 Early visit n= 222Control n= 232OR [95% CI]
  1. *Pap smears are performed on the majority of women at the first antenatal visit.

Urine test24(10.8)25(10.8)1.00 [0.561.81]
Blood pressure155(69.8)166(71.6)0.92 [0.61–1.38]
Weighed46 (20.7)56(24.1)0.82 [0.53–1.28]
Breast exam.49(22.1)55(23.7)0.91 [0.59–1.41]
Vaginal exam.52 (23.4)87 (37.5)0.51 [0.34–0.77]
Stitches check70(31.5)69 (29.7)1.09 [0.73–1.621
Pap smear*40(18.0)91 (39.2)0.34 [0.22–0.52]
Abdomen exam.170(76.6)163(70.3)1.38 [0.91–2.10]
Blood test17(7.7)14(6.0)1.29 [0.63–2.65]

There were no differences in the breastfeeding rates between the two groups at three and six months (Table 4). Survival analysis undertaken to look at the length of time spent breastfeeding (in weeks) showed no significant differences between the groups. There were no differences between intervention and control groups on any of the domains of the SF-36 at three or six months after birth. Table 5 gives the mean scores and standard deviations for intervention and control groups at three months.

Table 4.  Breastfeeding status for intervention and control groups at three and six months after birth. df = 1. Values are expressed as n(%).
 Early visit n= 232Control n= 243χ2P
Three months after birth
  Fully breastfeeding106(46.3)124(51.4)2.580.28
  Partial breastfeeding11(4.8)6(2.5)  
  Bottle feeding112(48.9)111(46.1)  
Six months after birth
  Fully breastfeeding63 (29.3)78 (34.5)1.570.46
  Partial breastfeeding18(8.4)20(8.9)  
  Bottles feeding134(62.3)128 (56.6)  
Table 5.  SF 36 and EPDS scores for early visit and control groups at three months after birth. Unless otherwise shown, values are expressed as mean [SD].
 Early visit n= 232Control n= 243P
SF 36 scores
  Physical functioning86.8 [17.8]86.5 [16.3]0.85
  Role-physical73.7 [35.7]72.1 [36.1]0.63
  Bodily pain77.8 [22.9]75.9 [23.1]0.39
  General Health74.4 [19.7]74.6 [19.010.93
  Vitality53.5 [20.1]53.1 [22.3]0.83
  Social functioning78.3 [24.0]79.4 [21.9]0.61
  Role-emotional76.2 [36.1]74.3 [38.5]0.57
  Mental health70.3 [19.7]72.1 [18.1]0.28
  Reported health transition3.06 [0.89]3.04 [0.86]0.8
EPDS score
  Mean score7.38 [5.31]7.48[5.35]0.85
  ≥1316.6%13.6%0.37

The percentage of women scoring as depressed (EPDS > 12) at three (intervention = 16.6%vs control = 13.6%, χ2= 0.8, df = 1, P= 0.37) or six (intervention = 11.6%vs control = 12.8%, χ2= 0.2, df = 1; P= 0.69) months postpartum did not differ significantly between the groups. There were also no differences in the mean scores at three (intervention = 7.38, sd = 5.31 vs control = 7.48, SD = 5.35; P= 0.85) or six (intervention = 5.87, SD = 5.37 vs control = 6.08, SD = 5.14; P= 0.67) months after birth. Mean EPDS scores and percentages of women scoring > 12 at three months postpartum are shown in Table 5.

There were no differences in the numbers of problems encountered in the six months following birth between intervention and control groups. Table 6 gives details for each problem listed in the three month survey and shows that women in the intervention group were more likely to report difficulties with low milk supply and adjusting to the demands of a new baby.

Table 6.  Problems encountered in the three months following birth. Denominators may vary due to missing values.
ProblemsEarly visit n= 232 (%)Control n= 243 (%)[95% CI] OR
Sore perineum72(31.0)62(25.5)1.31 [0.88–1.96]
Caesarean wound pain23 (9.9)28(11.5)0.85 [0.47–1.51]
Urinary incontinence20(8.6)17(7.0)1.25 [0.65–2.44]
Constipation/loss of control45(19.4)47(19.3)1.00 [0.64–1.58]
Haemorrhoids56(24.1)56(23.1)1.06 [0–69–1.62]
Tiredness & exhaustion131 (56.5)123(50.6)1.27 [0.88–1.82]
More cough/colds/minor illnesses35(15.1)25(10.3)1.55 [0.89–2.67]
Backache95 (41)103 (42.4)0.94 [0.66–1.36]
Mastitis27(11.6)38(15.6)0.71 [0.42–1.20]
Nipple pain64(27.6)62(25.5)1.11 [0.741.67]
Perceived low milk supply64(27.6)44(18.1)1.72 [1.12–2.66]
Feeling depressed or very unhappy for more than a few days43(18.5)38(15.6)1.23 [0.76–1.98]
Contraception84(36.2)85(35)1.06 [0.73–1 .54]
Sex45(19.4)50(20.6)0.93 [0.59–1 .45]
Relationship with partner33(14.2)39(16.1)0.87 [0.53–1 .43]
Losing weight51 (22)61 (25)0.84 [0.55–1.28]
Adjusting to demands of new baby61 (26.3)41 (16.9)1.76 [1.13–2.74]

Women were asked at three and six months whom they had talked to about caring for their baby and/or about their own health. Women in the intervention group were more likely to have talked to a GP (68.2%vs 58.0%, χ2= 5.33; P= 0.02) about caring for their baby and less likely to have talked to a hospital doctor (7.3%vs 14.0%, χ2= 5.6; P= 0.02) about caring for their baby in the three months following birth than women in the control group. These differences were not evident at six months.

Satisfaction with general practice care received in the six months following birth did not differ significantly between the two groups on either a 5-point global rating of care or on items dealing with particular aspects of care such as: ‘the GPs I visit under-stand the way I have been feeling since the birth’ and ‘the GPs I see are willing to spend time with me’, rated on 5–point Likert scales.

Between birth and six weeks women in the intervention group recorded a mean of 1.2 visits to their GP (95% CI 0.9–1.4) and women in the standard care group recorded a mean of 1.1 visits to their GP (95% CI 0.9–1.3), apart from the postnatal check-up visit.

DISCUSSION

A change in the timing of the routine postnatal check-up was the focus of this study. The early check-up gave general practitioners the opportunity to meet the mother and her new baby early in the postnatal phase. There was opportunity to focus on all aspects of maternal and child health; physical, emotional and social. General practitioners were not given detailed instructions or advice about what they should do at this check-up as the intervention was limited to changing the timing only.

The trial was successfully implemented at both centres, reaching the required sample size within the timeframe proposed. Response rates to follow up surveys were slightly lower than anticipated but gave enough power to test the prespecified hypotheses.

The intervention for the trial was an appointment time for an early check-up with a GP. As all participants gave informed consent it is possible that women allocated to the control group may have benefited from their involvement as participating in the trial made them more aware of their general practitioner in the early postnatal phase. Indeed, there was no difference between intervention and standard care groups in the number of visits to a GP in the first six weeks after childbirth. It is possible that both groups were given the message that to visit a GP early was OK and potentially of benefit. This is different from usual care when women are discharged from hospital with a letter to take to their GP six weeks later. There is no message that they should visit earlier. It may actually imply the reverse; you should be fine until six weeks. There is no way of knowing the magnitude or direction of this effect on the results of the trial but the possibility that it masked a significant finding should be remembered. The women who took part in this trial were all public patients, able to read and write English, and gave birth to a well baby.

The data were analysed according to allocation to the early or traditional group and not on whether the woman attended the check-up at the scheduled time. Calculation of the required sample size for the trial was based on being able to detect a 15% increase in breastfeeding rates at six months, a sample size which also allowed detection of clinically meaningful changes on the SF-36 questionnaire. The sample size obtained had 80% power (α= 0.05) to detect at least a 13% increase in breastfeeding from the six month mean of 35% (control group) and a shift in the mean score on the EPDS from 6.08 (control group, six month mean) to 4.6 and changes on the SF 36 of between 4.5 points for the physical functioning domain and 10 points for the role-emotional domain. A condition such as low back pain results in a score which is about 20 points lower than the general population on the physical functioning and role-emotional domains of the SF-3617.

There were no significant differences between control and intervention groups demonstrated on any of the major health outcomes studied. Women in the intervention group were significantly more likely to discuss labour and birth at their postnatal check-up. It is not possible to determine the exact nature of the discussion that took place. However, it has been shown that the opportunity to talk can result in health gains for women18. The fact that women in the intervention group were more likely to discuss their birth experience is not surprising considering their check-up was nearer the event. This is a subtle, but potentially powerful, change in the nature of the postnatal check-up.

Women in the intervention group were significantly less likely to attend for their postnatal check up at the scheduled time than women in the control group. This may have been because they saw the check-up as ‘new’ and therefore not important or the timing was not practical for some women. Perhaps knowing about the project made women feel confident to decide their own timing for the check-up as 96% of women reported attending a check-up at some time. A number of women commented on their surveys that the one week check was too early. Many stated this because they were not able to have an internal examination. It is not possible to say whether these women would have rated their early check-up differently if they realised an internal examination was not medically indicated. This raises an important issue which appears to be confusing both GPs and women.

Part of the rationale for trialing an early check-up was to provide a stimulus that might shift the focus of the check-up from a routine vaginal examination to an opportunity to spend time talking with a GP. This shift in focus has been recommended by a number of authors2,7,9,10 based on a study which showed that routine vaginal examinations provided little, if any, useful information8. The letter to the doctor at one week stated that a vaginal examination was not expected and should occur only if they felt it was medically indicated. However, a recent survey shows that some GPs still place importance on the routine vaginal examination as do some of the women in this study. These beliefs appear to have coloured the GP's interpretation and the woman's experience of the early check-up. The implication is that GPs would benefit from medical updates and practice guidelines for the content of the postnatal check-up and women would benefit from explanation about what is required at a postnatal check-up and why.

The intervention resulted in a change in the use of health services in the three months after birth. Women in the intervention group were more likely to talk to a GP about their baby's health than women in the control group and less likely to talk to a hospital doctor. There was no impact on visits to the maternal and child health nurse or other health professionals during this time. This finding suggests that women allocated an early visit developed more of a community focus in seeking health care for their baby (but not for themselves) than women in the control group, who were inclined to seek hospital care.

There were no differences in the number of problems reported by women in the three months following birth. However, women in the intervention group were more likely to report problems with low milk supply and adjusting to the demands of a new baby. These differences did not remain significant at six months yet they are interesting as they are both issues that are particularly relevant to the early postnatal weeks and likely to have become less so by six weeks. Perhaps women in the intervention group had the opportunity to talk about these issues with their GP thus ‘legitimising’ the concern and making them more likely to recall it at the time of the three month survey. There is also the possibility that an early visit undermined women's confidence by inferring they needed professional advice at early stage; low milk supply is a perception rather than a physiological reality. These are possibly chance findings but also hypotheses which should be incorporated into planning for other postnatal interventions.

Whatever the reason for the differences between the groups the finding illustrates how interventions can result in differences not anticipated in the original study design. It also illustrates the reason why trial sample sizes should be large enough to detect both improvements and deteriorations due to the intervention (ie, two-tailed). To make clinically important improvements in maternal health more is required than early postnatal review. The next step is to trial carefully planned, evidence based, postnatal educational interventions aimed at health professionals and women.

Acknowledgements

The authors would like to thank all the women who took part in the trial and members of the project reference group for their time and interest. J.G. was supported by a doctoral scholarship from the General Practice Evaluation Program, and J.L. is supported by a program grant from the Victorian Health Promotion Foundation. Project funding was obtained from the Royal Australian College of General Practitioners, Trainee Scholarship, Research and Development Fund.

Ancillary