Indication for Caesarean Delivery at the Princess Royal Maternity Hospital Glasgow; a study comparing the years 1962, 1992, 2002 and 2012
Eilidh Stewart, University of Glasgow
Introduction
There is long standing public health concern regarding the worldwide phenomenon of increasing caesarean section (CS) delivery rates. In 1985 the World Health Organisation (WHO) determined the optimal rate for delivery by CS to be no higher than 10-15% [1]. Despite this recommendation, rates continue to rise, particularly in developed countries. This includes our own national data in Scotland where in 1975 CS accounted for 8.6% of deliveries, increasing to 27.8% in 2011/12[2].
CS is a surgical procedure, which may be elective (ELCS) or emergency (EMCS), and is performed for various indications. However, there are specific associated risks which can affect both the mother and the child. Short term risks to the mother include infection, haemorrhage, thromboembolic events, and damage to abdominal organs including the GI and urinary tract. Long term effects on the mother impact primarily upon future pregnancy and include increased rates of placenta praevia, acreta and percreta, placental abruption, antepartum stillbirth and uterine rupture [3].
Additionally, for the infant there is a risk of laceration during the procedure along with an increased respiratory related morbidity risk [3]. Furthermore, the cost of CS poses significant implications; NICE calculated that the average cost of a planned CS in 2009/10 was £2369 compared to £1665 as the average cost for a planned vaginal birth [4].
As previously mentioned rates of CS in Scotland remain above that recommended by WHO, with a trend to increase further in the future [2]. It is therefore of interest to examine the documented clinical reasons for CS delivery with a view to developing strategies which may stop, or even reverse, this trend.
Glasgow specific data concerning indications for CS was last published in 1998, where Leitch et al performed a retrospective descriptive study looking at births in the Glasgow Royal Maternity Hospital in 1962 and 1992[5]. They found an increase in CS rates from 6.8% in 1962 to 18.1% in 1992; however they suggested that the increase was due to a lowering in the threshold of decision making, rather than a change in practice [5].
Since this initial study, the obstetric facilities for central and east Glasgow have amalgamated within a large general teaching hospital. The Princess Royal Maternity Hospital (PRMH) is a tertiary referral centre with a delivery rate of 6082 per year based on the 2010 national figures [2]. Additionally, several demographics factors of the population have changed since the initial study. In particular, it is accepted that mothers are getting older, which is demonstrated by the rise in the percentage of mothers 35years or over, from 6.0% in 1975/76 to 19.9% in 2011/12 [2]. Maternal smoking status has also changed, and the percentage of women who smoked at the time of their booking appointment in 2011/12 had dropped to 19.3%, compared to 25.4% in 2000/1 [2].
We therefore aimed to complete a follow up to the 1962 and 1992 data by performing a retrospective investigation of the indications for delivery by CS at the PRMH Glasgow during the years 2002 and 2012. Additionally, we aimed to examine maternal and delivery demographics, for all women attending for delivery at the PRMH Glasgow during the years 2002 and 2012.
Methods
The data search was conducted using the PROTOS database system which holds information, entered by midwives, concerning women giving birth at the PRMH. The required data was extracted and collated on an excel spreadsheet (Microsoft Excel 2007) by matching the patient ID number. The information on each patient collected included: method of delivery, the primary and secondary indications for CS, maternal age at the time of delivery, smoking status at the time of booking appointment, birth weight, and the estimate of gestation at the time of delivery. We excluded multiple births (twins/triplets) and non-livebirths in the index pregnancy (terminations of pregnancy, miscarriage and stillbirths).
Within our study, method of delivery was categorised in accordance with the original Leitch paper, spontaneous vertex delivery (SVD), instrumental vaginal delivery (IVD) and breech vaginal delivery (Br) [5]. CS deliveries were categorised into ELCS or EMCS, and where data had been entered using RCOG based categories 1-4, we classified category1-3 as EMCS and category 4 as ELCS.
The indications for CS were categorised into five groups as used by Leitch [5]; 1) previous CS, 2) failure to progress, 3) fetal indications, 4) malpresentation, and 5) “other”; including adverse obstetric history and maternal disease. Primary and secondary indications for CS were commonly recorded, and some were recorded freehand. These cases were considered on an individual basis and allocated to the most appropriate category.
Analysis was performed using SPSS, version 19 (SPSS Inc, Chicago, IL). Data are presented as means, SD, numbers and percentages as appropriate. Differences between groups were examined using ANOVA and two sample t-tests for continuous data and Chi-squared or binary logistic regression for categorical data. Odds ratios (OR), 95% confidence intervals (CI) and p values were calculated to compare between years for methods of delivery and the indications for CS, and were calculated using binary logistic regression analysis. Comparisons with 1962 or 1992 as a reference were calculated by hand as we did not have access to the raw data.
Before commencing the study, we confirmed that ethical approval was not required with Research and Development at Greater Glasgow and Clyde. However, we did gain Caldicott Guardian approval for the use of the data before commencing the study. Standard security measures were taken to protect data, which was stored by ID, with no patient identifiable information.
Results
Table 1 presents the demographic information of these women where we can see that the populations of women giving birth in 2002 and 2012 are significantly different in several respects. Average age of the women giving birth increased from 28.89 years to 29.34 years (p <0.001). Additionally, there was a decrease in the percentage of women giving birth who were smokers; 25.5% in 2002 to 11.5% in 2012 (p <0.001). The average length of gestation decreased from 39.20 to 39.10 weeks (95% CI (-0.18, -0.024), p <0.001). Birth weight of the babies increased over the ten year period by an average of 42.08 grams (95% CI (18.41, 65.75), p <0.001).
The method of delivery in the years studied is illustrated in Table 2. The rate of CS significantly increased from 22.94% in 2002 to 32.86% in 2012 (OR 1.64, p <0.001). Women were 6.59 times more likely to have a CS delivery in 2012 than in 1962 (p<0.001).
Delivery by SVD demonstrated a downward trend from 1962 to 2012 (76.19% vs 54.28% respectively), however, stayed relatively static over time as a percentage of non-CS deliveries undertaken (1962 vs 2012, 81.85% vs 80.85%). There was no significant change in breech deliveries between 2002 and 2012 (OR 0.60, p =0.29), however there was a large decrease in 2012 compared to 1962, (OR 0.05, p<0.001).
Data concerning indication for all CS deliveries is illustrated in Table 3, with EMCS and ELCS data in Table 4 and Table 5 respectively. Increases in both ELCS (7.0% vs13.3%, OR2.01, p<0.001) and EMCS (15.9% vs 19.5%, OR 1.28, p<0.001) were observed in 2012 compared to 2002.
Previous CS was the most common primary indication overall and specifically for ELCS, with a significant increase between years (51.0% vs 69.0%, OR2.14,p<0.001). Increases in the indication for EMCS were seen for fetal reasons (35.8% vs 44.9%, OR1.46, p<0.001), and previous CS (2.6% vs 7.3%, OR2.94, p<0.001), with a significant reduction in EMCS performed for FTP (36.6% vs 29.3%, OR0.72, p=0.001).
Discussion
In this retrospective, population based study we examined the indications for delivery by CS at the PRMH Glasgow during the years 2002 and 2012, and have been able to make comparisons with previously published data from a similar Glasgow based population in 1962 and 1992 [5].
The changes in maternal demographics are in agreement with national statistics [2]. The increase in the average age of women giving birth is believed to be influenced by several factors, and is a well documented risk factor for EMCS [2]. Our observed reduction in the numbers of women smoking between 2002 and 2012 could be influenced by the introduction of a public smoking ban in Scotland in 2006. Smoking is a well established risk factor for adverse pregnancy outcomes for mother and baby [2]. It is important to note the high percentage of “unknowns” (9.8% in 2012). This is higher than national data for 2011/12 which recorded 5.3% [2]. This indicates that smoking status in mothers is under-reported and therefore questions the quality and completeness of the data.
The increase in the total numbers of CS deliveries which were undertaken at PRMH is also in agreement with national and international data. The rate of CS at the PRMH in 2012 was found to be 32.86%, which is higher than the estimated Scottish average (27.8%) for the same year [2], but may be explained in part by the fact that the PRMH is a tertiary centre and therefore serves a higher risk population.
Interestingly, in our population, the percentage of all CS deliveries which were carried out as elective procedures increased from 2002 to 2012, now accounting for over 40% of CS deliveries and 13% of all deliveries. In addition, 2012 data indicates that previous CS is the most common primary indication, where previously failure to progress had been more common.
This data therefore suggests that whilst CS rates in our data set are high and increasing, this seems to be driven by an increase in elective procedures, with previous CS as the primary indication. This may be related to obstetric practice where women who have had a previous CS are routinely offered the option of repeat CS or vaginal birth after CS (VBAC) [4].
The method of delivery of pregnant women who have had a previous CS is therefore an important determinant of the overall CS rate. One suggested method to address the high number of CS is a dedicated antenatal counselling clinic for pregnant women who have had a previous CS. It is anticipated that this clinic would be developed to provide the opportunity to assess risk factors associated with VBAC and elective CS and also provide information to address concerns and allow expectant mothers to make a fully informed decision on their choice of delivery. In addition to information regarding current clinical practice, our data therefore provides baseline measurements to assess the success of this clinic.
Additionally, despite a smaller increase in EMCS rate, data suggest that over this ten year period, there may be a lower threshold for CS in the presence of suspected fetal problems with a concomitant reduction in EMCS performed for FTP. The significant increase in EMCS performed for fetal indications may indicate a lowering in threshold for CS or perhaps an increase in the numbers of women undergoing intrapartum fetal monitoring. This is of additional interest as since 2002, there has been in increased consultant presence on the labour ward which may have influenced decision making.
Conclusions
In conclusion, we have completed a population based follow up study, covering 50 years, examining mode of delivery and indication for CS at the PRMH Glasgow. Several demographic factors, which have changed within our population, reflect changes that have been observed at a national level.
We have demonstrated increases in total numbers of CS deliveries which seems to be driven by elective procedures carried out with the primary indication of previous CS. Additionally, despite a smaller increase in EMCS rate, data suggest that over this ten year period, there may be a lower threshold for CS in the presence of suspected fetal problems with a concomitant reduction in EMCS performed for FTP.
This provides essential information whereby obstetric clinicians and midwives can address current practices and performance against recommended UK practices. Although relative risks during CS delivery are small, there are recognised complications and financial costs associated with CS, and it is hoped this data can help to focus the development of methods with the aim of lowering the CS.
Word Count: 1992
References
1. The World Health Organisation. Appropriate Technology for Birth. Lancet 1985; 2(8452):436-7
2. Births in Scottish Hospitals, year ending 31 March 2012. Edinburgh: Information Services Division (ISD), NHS National Services, Scotland 2013.
3. Royal College of Obstetricians and Gynaecologists (2009) [Caesarean Section]. [Consent Advice No. 7]
4. National Institute for Health and Care Excellence (2011) [Caesarean Section]. [CG132] London: National Institute for Health and Care Excellence
5. Leitch CR, Walker JJ. The rise in caesarean section rate: the same indications but a lower threshold. BJOG: An International Journal of Obstetrics and Gynaecology 1998: 105(6):621-26
Tables
Year of Delivery
2002 / 2012 / p
Births / N / 4101 / 6037
Age (years) / Mean (SD) / 28.89 / (6.16) / 29.34 / (5.79) / <0.001 / ***
Gestation (weeks) / Mean (SD) / 39.20 / (1.99) / 39.10 / (1.95) / 0.01 / **
Birthweight (grams) / Mean (SD) / 3336 / (616) / 3378 / (580) / <0.001 / ***
SIMD Quintile / 1 / N (%) / 1740 / (42.8) / 2876 / (47.8) / <0.001 / ***
2 / 678 / (17.5) / 1176 / (19.6)
3 / 559 / (14.4) / 799 / (13.3)
4 / 442 / (11.4) / 654 / (10.9)
5 / 464 / (11.9) / 508 / (8.4)
Unknown / 218 / (2.1) / 24 / (0.3)
Smoking Status / Non-S / N (%) / 2840 / (74.3) / 4752 / (78.8) / <0.001 / ***
Smoker / 974 / (25.5) / 691 / (11.5)
Unknown / 6 / (0.2) / 591 / (9.8)
Gender / Male / N (%) / 2051 / (50.0) / 3026 / (50.1) / 0.706
Female / 2050 / (50.0) / 3010 / (49.9)
*=p≤0.05, **=p≤0.01, ***=p≤0.001