Compliance with routine antenatal rhesus D prophylaxis and the impact on sensitisations: observations over 14 years
Dr IZ MacKenzie, Nuffield Department of Obstetrics and Gynaecology, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK. Email firstname.lastname@example.org
Documented routine antenatal anti-D prophylaxis was given to 90% and 81–87% of eligible women at 28 and 34 weeks of gestation, respectively, during the early 1990s and early 2000s. With increasing experience and education, a significant improvement in the timing of the first (OR 0.26, 95% CI 0.16–0.41: P < 0.0001) and second injections (OR 0.40, 95% CI 0.26–0.61: P < 0.0001) occurred during the latter period. Despite these improvements, there was no reduction in the sensitisation rate at 0.4%. However, this low rate occurred despite significant proportions of women delivering more than 42 days after the second injection. Fifteen of the 16 sensitised women had received routine antenatal prophylaxis.
Routine antenatal rhesus (Rh) prophylaxis to reduce the chances of unprovoked sensitisation occurring during pregnancy was first proposed in the 1970s and has been a subject to investigation since that time. There is convincing evidence from many retrospective and a few prospective studies that it reduces the incidence of sensitisation from 1.5–2.0% to 0.1–0.4%, and the consensus conference on anti-D prophylaxis in 1997 recommended adopting routine antenatal prophylaxis.1
Our previous prospective study of routine antenatal prophylaxis, introduced in 1986, found relatively poor compliance.2 Despite this, the expected reduction in sensitisations was observed and it was unclear whether an even greater reduction could be expected with improved compliance.
To determine whether compliance has improved with increased experience and acceptance of antenatal prophylaxis together with increased publicity and education of the benefits and to discover if the sensitisation could be reduced further, the analysis has been extended.
An analysis examining the success in providing routine antenatal prophylaxis to RhD negative nulliparae delivering in the John Radcliffe Hospital and surrounding district during 1992–2003 has been conducted by scrutinising clinical records. The previous study2 explored the success during 1992–96 by examining the notes of every fifth consecutive delivery and these have been reviewed, re-analysed and are presented. The analysis has been extended to include all RhD negative women who delivered during 1997–2003; eligible women were identified from prospectively collected data used to generate the unit statistics.3 To identify a 10% improvement in compliance over the results in providing the first injection reported for the earlier period, power calculations indicated that the results of 160 women eligible for routine antenatal prophylaxis needed to be analysed to provide an alpha value of 0.05 with 80% power. From the experience of the earlier study, we reasoned that an initial identification of 240 women would provide sufficient numbers to allow for those who delivered before 28 weeks of gestation, those who booked later than 34 weeks of gestation and women found ineligible for other reasons. The obstetric notes of one in ten RhD negative nulliparae were scrutinised to look for evidence that routine antenatal anti-D prophylaxis had been administered. Note was also made of the administration of prophylaxis at other times during the antenatal period and at the time of delivery. The records maintained by the serology laboratory of the Oxford National Blood Service were interrogated to cross-check that the immunoglobulin vial had been issued for the named woman whenever evidence could not be found in the obstetric notes.
At all times, anti-D immunoglobulin G (IgG) doses of 500 iu (Bio Products Laboratory, Elstree, Herts, UK) were given at 28 and 34 weeks of gestation apart from January 1998–July 2002, when the 500-iu vials were withdrawn from clinical use; during this period anti-D IgG 1250 iu (Baxter Healthcare Ltd, Newbury, Berks, UK) was given at 28 and 34 weeks of gestation.
All RhD negative women with evidence of sensitisation during their second pregnancy who delivered in the John Radcliffe Hospital and surrounding district between 1 January 1990 and 31 December 2003 were identified through the Oxford Rhesus Therapy Unit database and the Antenatal Serology Laboratory records in the Oxford National Blood Service. Details of the women’s previous pregnancy management with particular regard to where they were confined and whether routine antenatal prophylaxis and other prophylaxis injections had been given were noted. The number of RhD negative women and their parity were obtained from the unit statistics, since parity and Rh type are routinely recorded; for our previous study for 1990–96, the denominators for these had been calculated. Statistical analyses were performed using odds ratios and 95% confidence intervals where appropriate.
For the investigation into compliance, 365 of 480 case records selected as every fifth delivery in 1992–96 and 215 of 240 selected as every tenth consecutive delivery in 1997–2003 were reviewed. Women were excluded if the records were not obtainable, the women had been involved in a specific routine antenatal prophylaxis study or the parity or Rh type had been incorrectly coded.
Table 1 illustrates the results for both time periods for both injections. Three of the 340 women eligible for prophylaxis during 1992–96 refused prophylaxis on one or both occasions compared with seven during 1997–2003 (OR 0.23, 95% CI 0.06–0.91: P < 0.04).
Table 1. Documented evidence that prophylaxis had been given
|Study period||1992–96||1997–2003|| ||1992–96||1997–2003|| |
|Women‘s records examined||365||215|| ||365||215|| |
|Eligible for prophylaxis*||350||196|| ||340||197|| |
|Refused prophylaxis||3||5||0.33 (0.08–1.40)||3||7||0.23 (0.06–0.91), P < 0.04|
|No. available for prophylaxis||347||191|| ||337||190|| |
|Received injections||309 (89%)||171 (90%)||0.95 (0.08–1.40)||273 (81%)||165 (87%)||0.65 (0.39–1.07)|
|At correct gestation (± 1 week)||183 (59%)||145 (85%)||0.26 (0.16–0.42), P < 0.0001|| |
|At earlier gestation||37 (12%)||4 (2%)||5.68 (1.99–16.22), P < 0.0001|| |
|At later gestation||89 (29%)||22 (13%)||2.74 (1.64–4.57), P < 0.0001|| |
|Received both injections|| ||251 (74%)||150 (79%)||0.78 (0.51–1.19)|
| 5.5–6.5 weeks|| ||125 (50%)||107 (71%)||0.40 (0.26–0.61), P < 0.0001|
| <5.5 weeks|| ||32 (13%)||9 (6%)||2.29 (1.06–4.94), P < 0.04|
| >6.5 weeks|| ||94 (37%)||34 (23%)||2.04 (1.29–3.24), P < 0.003|
During 1992–96, 309 of 347 ‘available’ women (89%) received the first injection and for 183 (53%), the injection was given within a week either side of 28 weeks of gestation. During 1997–2003, 171 of 191 ‘available’ women (90%) received the injection and for 145 (85%), it was given at the correct gestation. The improvement in the timing of the first injection during the second observation period was significant (OR 2.17, 95% CI 1.45–3.24: P < 0.0001).
During 1992–96, 273 of 337 ‘available’ women (81%) received the second injection. Of the 251 women who received both prophylaxis injections, 125 (50%) received the second injection after an interval of 5.5–6.5 weeks. During 1997–2003, 165 of 190 ‘available’ women (87%) were given the second injection. Of the 150 women who were given both injections, 107 (71%) received the injection after an interval of 5.5–6.5 weeks. There was no significant improvement in the proportion of women given the second injection during the two study periods, but there was a marked improvement in the timing between the two injections during the latter period (OR 0.40, 95% CI 0.26–0.61: P < 0.0001).
Second injection to delivery interval
During 1992–96, 87 of 273 (32%) women given the second injection delivered at 40 weeks of gestation or later compared with 51 of the 165 (31%) women delivered during 1997–2003. During the first study period, 18% delivered more than 6 weeks after the second injection, with 6% delivering more than 7 weeks later, while the figures for the second study period were 19% and 7%, respectively.
Additional anti-D prophylaxis was given during the antenatal period to 30 women during 1992–96; 8 followed an amniocentesis, 21 after an episode of vaginal bleeding, while in 1, the reason was not discernible. The respective figures for 1997–2003 were 19 in total; 6 after amniocentesis, 9 after vaginal bleeding, 1 following external cephalic version and 3 where the reason was not discernible.
Investigation into postdelivery anti-D prophylaxis was not conducted for the period 1992–96. During 1997–2003, postdelivery injections had been documented for 202 of the 215 (94%) women; in five cases, the injection was refused (antenatal prophylaxis was also refused) and in eight there was no record that we could find in the clinical notes that the postdelivery injection had been given. In four women, the delivery had occurred before 28 weeks of gestation, in two it was at 33 and 35 weeks of gestation, respectively, and in two it was at term; appropriate antenatal prophylaxis injections had been given to these latter four women. In 1992–96, where the baby’s Rh type had been recorded in the obstetric notes, 63% of them were Rh (D) positive compared with 62% in 1997–2003.
A total of 14 500 Rh negative women were delivered during the study period analysing sensitisation rates; the relevant populations for the periods 1990–96 and 1997–2003 are shown in Table 2. Included are the total numbers of para 1 delivered and a calculation for those who delivered their first baby in the Oxford district, allowing for the 16% of women who we had previously shown had their first confinement in another district. Since information on routine antenatal prophylaxis was not available, they have been excluded from the analyses.2
Table 2. Populations studied, sensitised women and sensitisation rates for 1990–96 and 1997–2003
|Total women delivered||45 997||43 484||89 481|
|Total Rh negative women delivered||7538||6962||14 500|
|Total Rh negative nulliparae||3182||3084||6266|
|Total Rh negative para 1 delivered||2685||2440||5125|
|Total sensitised women||16||12||28|
|Total sensitisation rate||0.60%||0.49%||0.55%|
|Rh negative para 1 stable Oxford population*||2255||2050||4305|
|Sensitised women, non-Oxford||8||8||16|
|Rh negative para 1 ‘immigrant’ population**||430||390||820|
|Sensitised women, non-Oxford||8||4||12|
There were 16 para 1 delivered during 1990–96 who were RhD sensitised; eight of these women had delivered their first baby in Oxford, seven had received both routine antenatal injections and one woman received only one injection at 36 weeks of gestation. Three of the women delivered beyond their expected date of confinement; all received routine postdelivery prophylaxis. One woman showed evidence of active sensitisation by the end of the pregnancy and one probably experienced a large fetomaternal haemorrhage at 35 weeks of gestation and delivered a very anaemic baby at 37 weeks of gestation.
There were 12 para 1 delivered during 1997–2003 who were RhD sensitised; eight of these women had delivered their first baby in the Oxford district between 1991 and 1998 and four had delivered their first baby outside Oxford. Seven of the eight delivered in Oxford had been given routine antenatal prophylaxis at the appropriate times (for one woman there was documented evidence that she only received the 28-week injection) and three of the seven had evidence of active sensitisation by the end of the first pregnancy; postdelivery prophylaxis was given appropriately to six of them and only one woman delivered beyond her expected date of confinement. The obstetric notes of the final woman, who had a needle phobia and whose first baby was delivered in 1998, contained no record that antenatal prophylaxis had been given and there was no evidence of sensitisation by the end of the pregnancy.
Also shown in Table 2 are the numbers of Rh (D) negative para 1 delivered during 1990–96 and 1997–2003 in the Oxford district and the numbers with a sensitised pregnancy according to where the first baby was delivered. The sensitisation rate for ‘Oxford’ women for the two periods remained constant at 0.3–0.4%. The sensitisation rate for the ‘non-Oxford’ women reduced from 1.6 to 1.3%, a statistically nonsignificant reduction (OR 1.27, 95% CI 0.40–4.02); this reduction could be due to routine antenatal prophylaxis having been given to some of these women. The sensitisation rate for ‘Oxford’ women was, however, significantly lower than for the ‘non-Oxford’ women (OR 0.30, 95% CI 0.10–0.93).
There was a reassuring consistency in the results obtained between the two observation periods, with regard to the proportion of women who delivered beyond their expected dates of confinement, the proportion who received additional antenatal prophylaxis for sensitising events and the neonatal Rh type.
There was an improvement in providing routine antenatal prophylaxis during the later period, with a significantly greater proportion receiving the 28-week injection at the correct gestation and significantly more women receiving the second injection after a 6-week interval. However, there remained approximately 10% of cases where there was no documentary evidence that the 28-week gestation injection had been given and 13–19% of cases with no documentary evidence that the 34-week gestation injection had been given. We acknowledge that the injections may have been given but not documented. During the past 10 years, there has been a marked improvement in documentation in obstetric case records and the relatively low-documented administration rates recorded during the earlier period possibly reflected failed documentation rather than failed drug administration.
The proportions of eligible women who were available to receive prophylaxis among the women whose case records were reviewed remained at a relatively constant rate during the two time periods. It is of note, however, that 3 of 350 (0.8%) women refused to have at least one antenatal prophylaxis injection during the early 1990s and that the proportion had increased four-fold to 7 of 197 (3.5%) by the end of the 1990s. There has been a growing anxiety about possible infection from the administration of blood products during the decade, and such anxiety may well have been exacerbated when the preparation previously used for Rh prophylaxis was withdrawn because of concerns relating to variant Creutzfeldt–Jakob disease transmission. Given the clear benefits produced by anti-D prophylaxis, the impetus to develop a recombinant antibody for protection against Rh D sensitisation continues to grow to try to ameliorate this anxiety.
The issue of the duration of the therapeutic benefit of antenatal injections remains relevant if the effective period is limited to 42 days after the prophylactic injection has been given. Delivery occurred more than 42 days after the second injection in approaching 20% of women. These women were thus theoretically unprotected against sensitisation. Recent evidence examining the duration of detectable passive anti-D in the maternal circulation has shown that for many women, the decline in detectable antibody occurs much sooner than 42 days.2,4,5 Although the use of the two-dose schedule employed in this study has a theoretical advantage over the single 28-week dose schedule, laboratory studies do not confirm any clear advantage. However, the single-dose strategy eliminates the risk of women not receiving the second dose in the two-dose strategy (13% in the later period studied) or an inappropriately long interval occurring between the two doses (23%).
Despite the reservations with regard to the one- or two-dose strategy and the lack of protection for women delivering beyond their expected date of confinement,2,4,5 the results obtained from the analysis of sensitised pregnancies strongly suggest that the critical protective level of passive anti-D in the maternal circulation is lower than had previously been thought necessary.6 Only 4 of the 16 sensitised women delivered beyond their expected date of confinement; one of them had received only one antenatal injection at 36 weeks of gestation and delivering at 41 weeks of gestation and another was the woman with no record of any antenatal injections who delivered at 42 weeks of gestation. With a high proportion of women delivering beyond their expected date of confinement, it would have been expected that many of the sensitised cases would have occurred in that group in which apparently inadequate antibody remained in the circulation by that stage, but this was not the case.
The explanation for the ‘failures’ remains to be identified since 14 of the 16 ‘Oxford’ women had been given both antenatal injections at the appropriate time; failure was possibly due to an inadequate uptake of antibody from the intramuscular deposition, as has been shown by recent serial longitudinal studies.4,5,7 Alternatively, an unprovoked fetomaternal haemorrhage causing sensibilisation may have occurred before 28 weeks of gestation, or an unrecognised large fetomaternal haemorrhage for which there was insufficient passive antibody to neutralise the RhD antigen may have been the explanation. Nevertheless, this study provides further evidence in support of routine antenatal prophylaxis and that the strategy involving a single injection at 28 weeks of gestation offers the expectation of improved compliance without compromising efficacy.
The authors wish to acknowledge the help provided by Ms Helen Loynes, data analyst at the John Radcliffe Hospital Maternity Unit and Ms Janet Scott for interrogating the database of the National Blood Service, Oxford.
The Oxford Rhesus Therapy Unit Trust Fund provided funding for hand-searching the obstetric notes.
Bio Products Laboratory and ZLB Behring have previously funded studies on antenatal anti-D prophylaxis in the unit.