Reactive thrombocytosis after caesarean section and vaginal delivery: implications for maternal thromboembolism and its prevention

Authors


Correspondence: Dr R. Atalla, Department of Obstetrics and Gynaecology, Queen Elizabeth II Hospital, Howlands, Welwyn Garden City, Hertfordshire AL7 4HQ, UK.

Abstract

Objective To assess the duration and severity of reactive thrombocytosis after caesarean section and vaginal delivery.

Design A prospective cohort study.

Setting A large teaching hospital.

Methods Women admitted for delivery at the Leicester Royal Infirmary were recruited into the study. The platelet count was measured before delivery and postnatally on days 3, 8, 12, 16, 20 and 24. Women who had antepartum haemorrhage, postpartum haemorrhage and those delivered by instrumental delivery were excluded from the study. Sixty-five were recruited, and 45 completed the study, 20 of whom were delivered by a normal vaginal delivery and 25 by caesarean section. A random effects model was used to compare platelet counts within and between the two groups to assess the severity and the timing of reactive thrombocytosis.

Results There were no statistically significant differences in booking and pre-delivery platelet counts between the two groups (mean values 248.4 × 109/L and 245 × 109/L in the normal vaginal group and 269.4 × 109/L and 251.6 × 109/L in the caesarean section group, respectively). Postnatally, a rise in the platelet count was noted in the normal vaginal delivery group, reaching statistically significant peak values, compared with booking and pre-delivery at days 8 and 12 of the postnatal period (mean value 365.8 × 109/L; P < 0.001 and 369.4 × 109/L; P < 0.001 respectively). In the caesarean section group, the platelet count was raised to a statistically significant high value, compared with booking and pre-delivery at day 8 of the postnatal period. The platelet count peaked at days 12 and 16 of the postnatal period (mean value 522.5 × 109/L; P < 0.0001 and 526.5 × 109/L; P < 0.0001, respectively) and remained significantly higher than booking and predelivery values for 24 days after the caesarean section. There was a greater rise in the platelet count in the caesarean section group compared with the vaginal delivery group. The platelet counts in the caesarean section group were significantly higher than these in the normal vaginal delivery group from day 12 to day 24 of the postnatal period.

Conclusion A significant rise in platelet count occurred eight to twelve days after normal vaginal delivery and caesarean section. The increase in platelet count continued to rise for 16 days after caesarean section, and it stayed significantly higher for more than 24 days after the delivery.

INTRODUCTION

Post-operative reactive thrombocytosis, which has been implicated in the genesis of thromboembolism, has been demonstrated after a variety of surgical procedures1–3. Acute blood loss during the operation, chronic blood loss post-operatively and anaemia are known to significantly increase the post-operative platelet count4–10. As these factors are frequently associated with caesarean section, we investigated the presence and severity of thrombocytosis following uncomplicated caesarean section and normal vaginal delivery.

METHODS

Women admitted consecutively to Leicester Royal Infirmary for delivery were recruited to the study. Those with proteinuric hypertension or medical disorders affecting the platelet count such as thrombocytopenia, were excluded from the study. After written consent was obtained, a pre-delivery full blood count was performed to assess the platelet count and haemoglobin concentration. Labour was managed according to unit guidelines. Women who were delivered by instrumental delivery and those who suffered from antepartum or postpartum haemorrhage of an estimated blood loss > 1000 mL were excluded from the study. Women who were delivered by caesarean section were assessed at time of surgery for their risk of developing thromboembolic complications according to guidelines produced by the Royal College of Obstetricians and Gynaecologists11. Women at moderate or high risk received unfractionated heparin in a dose of 5000 IU subcutaneously every 12 hours for five days or until fully mobile11. If postnatal infection was suspected or diagnosed, the woman was excluded from the study. A platelet count was performed on the third postnatal day after which the women were usually discharged home. A community midwife visited the women at their homes to take blood samples for platelet count on days 8, 12, 16, 20 and 24 and to confirm the absence of complications. The platelet counts were analysed using a random effects model to allow for the correlation between repeated measures12. This study was approved by the regional ethical committee.

RESULTS

Sixty-five women were included in the study. Of these, seven had an instrumental delivery, one suffered a post-partum haemorrhage and one developed a urinary tract infection during the postnatal period. Eleven others opted out of the study after they had been discharged from the hospital. Forty-five women completed the study, 20 of whom had a normal vaginal delivery and 25 of whom had an uncomplicated caesarean section (either elective or emergency). There were no statistical significant differences in the maternal age, booking weight or gestational age at delivery between the two groups (Table 1). In the caesarean section group, six women were delivered by emergency caesarean section and were considered to be of moderate risk of developing deep venous thrombosis. Post-operatively, they were started on unfractionated heparin in a dose of 5000 IU subcutaneously every 12 hours. The heparin was stopped on the third post-operative day as the women were fully mobile. Booking and pre-delivery platelet counts of all the women were within the normal range (150–400 × 109/L), and there were no statistical significant differences between those in the normal vaginal delivery group and those in the caesarean section group. There was a slight fall in the pre-delivery platelet count in both groups compared with booking, but this fall was not statistically significant (Table 2).

Table 1.  Characteristics of women in both study groups. Values are given as mean (range).
 Normal vaginal delivery (n= 20)Caesarean section (n= 25)
Maternal age (years)29.8 (22–38)30.8 (19–38)
Maternal weight (kg)68 (51–94)65 (47–88)
Gestational age (weeks)38.1 (34–41)39.1 (3442)
Table 2.  The mean (range) platelet count (× 109/L) in the two study
 Normal vaginal deliveryPCaesarean sectionP
  1. *Compared with booking and pre-delivery platelet count.

  2. **Compared with normal vaginal delivery group.

Booking248.4(186–338)269.4 (138–338)
Pre-delivery245 (161–331)251.6(146–368)
Day 3239.9(116–439)275.9 (145–444)
Day 8367.8(274–641)< 0–001*449.4 (229–615)< 0.001*
Day 12369.4 (207–559)< 0–001*522.5 (283–972)< 0.0001**
Day 16311.6(221431)526.5 (245–1217)< 0.0001**
Day 20306.1 (213–379)459.6 (227–667)< 0.0001**
Day 24275.9 (169–372)413.8 (119–757)< 0.0001**

In the normal vaginal delivery group, the platelet count continued to fall until the third postnatal day, but this decrease in platelet count was not statistically significant. The platelet count increased rapidly afterwards to reach a peak value between days 8 and 12 of the postnatal period when the platelets counts were significantly higher than booking and pre-delivery (P < 0.001) (Table 2). The platelet count exceeded the normal range (400 × 109/L) in only six women (30%). The mean platelet count decreased gradually thereafter.

In the caesarean section group, reactive thrombocytosis began on the third post-operative day. The platelet count gradually increased reaching a statistically significant value on the 8th post-operative day compared with booking and pre-delivery count (P < 0.0001). Platelet count continued to rise for a longer period, reaching a peak value between days 12 and 16 of the postnatal period and remained at a statistically significant higher level, compared with the booking count—and pre-delivery—until after day 24 of the postnatal period (P < 0.0001) (Fig. 1). The platelet count exceeded the normal range (400 × 109/L) in 20 women (80%). The platelet counts in women in the caesarean section group was significantly higher than in women in the normal vaginal delivery group from day 12 to day 24 of the postnatal period.

Figure 1.

Mean and one standard deviation of platelet count in the normal vaginal delivery group inline image and in the caesarean section group inline image.

DISCUSSION

Reactive thrombocytosis has been shown to occur after a variety of surgical and orthopaedic operations which are associated with a high incidence of post-operative thromboembolic complications. The severity of reactive thrombocytosis has been found to vary with the type of procedure performed1–3. However, reactive thrombocytosis and its relation to thrombosis have not been assessed systematically after caesarean sections and normal vaginal deliveries.

This study has demonstrated the occurrence of reactive thrombocytosis following caesarean sections as well as after uncomplicated normal vaginal deliveries. Reactive thrombocytosis was more prominent in the former, reaching a statistically significant high value on the eighth post-operative day, and continuing to rise to a higher peak value between days 12 and 16 postnatally. The platelet counts remained at a significantly high value for more than 24 days after the operation. These significantly high platelet counts were noted after uncomplicated caesarean sections. An even greater rise in platelet count would be expected in women with postnatal complications, such as postpartum haemorrhage or infection, as these conditions have been shown to exacerbate reactive thrombocytosis3–10.

Reactive thrombocytosis has been shown to be associated with an increased incidence of thrombosis. In the previously published small studies, 3%–11% of patients with a platelet count of between 400–1000 × 109/L due to post-operative reactive thrombocytosis developed symptomatic deep venous thrombosis13,14. This could partly explain why the risk of puerperal thromboembolic events is 10 times greater following caesarean section than normal delivery15–17 and are more common 10 days or more after the delivery11,18,19.

Over the last decade, the triennial reports of the confidential enquiries into maternal death showed that thromboembolic events were the main cause of maternal deaths in the United Kingdom18, with a mortality rate of 1.3 per 100,000 maternities (22.8%–27.1% of all maternal mortalities). This mortality rate has even increased in the last triennial report to 2.1 per 100,000 maternities (35.8% of all maternal mortalities)19. Of these deaths, 62% to 80% occurred in the postnatal period and over 80% of these occurred 10 days or more after the delivery day18,19; 60% to 76% of these women were delivered by a caesarean section18,19. Though, caesarean section was highlighted as the main risk factor for thromboem-bolism, the report emphasised that anaemia, infection and postpartum haemorrhage, which are known to cause reactive thrombocytosis4–10, were also additional risk factors11,18.

The use of an antiplatelet agent for at least two weeks after surgical procedures has been shown to reduce significantly mortality and morbidity due to post-operative thromboembolism and confers additional protective effects even when heparin is being used for the immediate post-operative period20. This adds weight to the clinical significance of reactive thrombocytosis and its severity in the development of thromboembolic complications in the postnatal period.

The occurrence of overt thromboembolism is perceived to be rare and does not reflect the true incidence of post-operative thrombosis as venous thromboembolism is most often clinically silent and many of the thrombotic events occur after discharge from hospital11,18,19. Of the episodes of pulmonary embolism, 70% to 80% are clinically silent with the diagnosis being made at autopsy21,22. Most patients will die within 30 minutes of the acute event, providing inadequate time for any anticoagulant therapy to be effective. Therefore, reliance on the diagnosis and treatment of established venous thromboembolism may expose susceptible patients to unacceptable risks when the first manifestation of the disease may be a fatal pulmonary embolism23,24. As about 80% of the postpartum deaths occurred 10 days or more following the delivery, more studies of prophylactic measures beyond the hospital stay are needed in order to reduce mortality and morbidity11,18,19. Prophylaxis against the risk of thromboembolic complications associated with reactive thrombocytosis would necessitate an antiplatelet agent that is simple to administer at home and relatively safe to use for about four weeks in the postnatal period. The efficacy of such a method for thromboprophylaxis was established after general and orthopaedic surgery20. Our findings of thrombocytosis after caesarean section add weight to the need for such a trial after caesarean sections.

Ancillary