Changing Trends of Obstetric Hysterectomy Over Last 2 Decades: A Retrospective Analysis from a Single Center Tertiary Care Rural Teaching Hospital of Western India

Abstract

Objective

The primary objective of our study to analyze the trend of obstetric hysterectomy (OH) over last 2 decade in rural tertiary care referral center of western India.

Design

Retrospective observational and comparative study from single rural tertiary care referral center of western India.

Method

A retrospective data collected of patients who underwent OH from Jan 2001 to Dec 2021. Each patient studied in detail for demographic and clinical profile, indication of OH, Intra operative and post operative complication and feto-maternal outcome. Patient of first decade (2001 to 2010) compared with second decade (2011 to 2021) to see the changing trend of OH.

Results

Total 19,666 patients delivered in study period.195 patient underwent OH. Incidence of OH is 0.9%. 171 out of 195 patients were referred (87%). If we see the trend of OH over 2 decades 2001 to 2010 and 2011 to 2021 incidence of OH slightly raised from 0.87% to 1.07% (p 0.16). Although Incidence of OH in cesarean delivery rises from 0.6 to 1.4% (p 0.6) but this change due to overall increase in cesarean section rate from 37 to 49%(p < 0.0001). Indication of OH significantly change as uterine rupture decrease significantly from 61 to 22%(p < 0.001) and morbidly adherent placenta increases from 7 to 23%(p 0.007). ICU admission and ventilator support increases from 54 to 79%(p 0.04) and 22% to 70% (p < 0.001), respectively, due to improved critical care services, but maternal death remains relatively same 15% & 20% (p 0.5).

Conclusion

Primary CS rates should be decreased to prevent adherent placenta spectrum disorders. Good antenatal care and development of a robust referral system can decrease maternal mortality. Greater utilization of skill laboratory and simulators will go a long way in developing the skills of aspiring doctors in operative vaginal deliveries.

Introduction

The dreaded triad responsible for the majority of cases of maternal mortality constitutes hypertension, infectious etiologies and obstetric hemorrhage of which postpartum hemorrhage is the leading cause. PPH is defined as > 500 ml and > 1000 ml blood loss after vaginal delivery and cesarean section (CS), respectively. PPH prevalence varies depending on the type of delivery: 4.9% for vaginal births, 18.4% for instrumental births, 8.5% for elective CS, and 19.8% for emergency CS [1]. In developing countries like India, nutritional deficiencies like anemia exacerbate PPH leading to a 38% stake in the total mortality in pregnant women (SRS 2001–2003), rural settings pose an additional challenge in delivering timely life-saving critical care such as Obstetric hysterectomies (OH). It serves as a life-saving intervention in cases where conventional conservative therapies have failed. The first obstetric hysterectomy during cesarean section was done by Porrow E in 1876 as a means to control PPH and prevent infection [3]. The last 2 decades have witnessed a shift in the course of modern obstetrics with a significant rise in CS rate projected at 28.5% worldwide by 2030 and 63.4% in eastern Asia [4].

The aim of our study was to identify the incidence, clinical profile, demographic characteristics, indications and outcomes of patients who underwent OH in the last 2 decades at our institute along with an analysis of trends of OH, as a result of evolved obstetric practices.

Material and Methods

We conducted a retrospective analytic and comparative study from January 2001 to December 2021. We defined Obstetric hysterectomy as a hysterectomy performed during or within 42 days of termination of pregnancy. All patients who underwent OH during the aforementioned period of time at Shree Krishna Hospital, Karamsad were included in the study. Shree Krishna hospital is a teaching institute and tertiary care referral center catering to urban, semi-urban and rural areas within a 50 km radius. Referral cases of women who underwent OH in another hospital were not included in this study. Significant data were extracted from the birth registers and the medical record department. Each patient who underwent OH was studied in detail for her clinical and demographic profile, indications, need for massive blood transfusion and maternal–fetal outcome. The data sets were divided into two groups, from January 2001 to December 2010 (the 1st decade) and January 2011 to December 2021(the 2 decade). A comparative analysis of both groups was done for different parameters. Hospital ethical review committee approval was taken to conduct the study.

Statistical Analysis

Data were entered into a Microsoft Excel spreadsheet and analyzed with the help of the SPSS 28.0.1 software. Descriptive statistics were used for the clinical and demographic profiles. Due to the extremes of range present within the sample size, median with interquartile range (Q1–Q3) was used. To compare variations in incidence rates, the incidence rate ratio was utilized. To compare the proportions of the two groups and analyzed qualitative data, chi-square and Fisher exact tests were performed. A p value of < 0.05 was considered significant to reject the null hypothesis.

Results

Of the total 19,666 deliveries performed in the study period, 195 of them were OH. The overall incidence of OH was 9/1000 livebirths. The rate of OH with CS was 15/1000 while that with vaginal deliveries was 3/1000. Along with an increase in overall number of deliveries in our facility (8025 to 11,641), the CS rate also increased significantly in the 2nd decade (37.3% to 49.13%, p value < 0.0001) (Table 1). In the 2nd decade, the incidence of OH increased to 10.7/1000 live birth from 8.7/1000 live births in the 1st decade; there was a reduction in the occurrence of OH among those who underwent CS (6 vs 14 per 1000 live births, p value 0.04).

Table 1.

Decade wise OH distribution with mode of delivery

Decade	No of Total delivery	No of vaginal deliveries	No of CS(CS rate)	No of OH	OH incidence/Live births
2001–2010(1st decade)	8025	5026(62.6%)	2999(37.3%)	70	8.7/1000
2011–2021 (2 decade)	11,641	5921(50.8%)	5720(49.3%)	125	10.7/1000
Total	19,666	10,947(55.6%)	8719(44.33%)	195	9/1000

The majority of the study group was made up of patients in the 21–30 years age group (70%), Multigravida (67.6%) and referred from other centers (87.6%) with a median parity of 2 (IQR 1–3) and the median age of 28 years (IQR 21–35 years) (Table 2). A complete hysterectomy was performed on 48.7% of patients, whereas a subtotal hysterectomy was performed on 51.2%. The occurrence of OH among primigravida reduced. (34%Vs15%. p value, 0.007).

Table 2.

Demographic parameters of patients

Demographic parameters	No of patient (n = 195)
Age
< 20 yrs	7(3.5%)
21-30yrs	135(69.2%)
> 30 yrs	53(27.17%)
Parity
Primi	43(22.05%)
Multi	132(67.6%)
Grand multi(> 5)	20(10.2%)
Booking status
Booked	26(13.3%)
Referred	171(87.6%)
Type of hysterectomy
Total	95(48.7%)
Subtotal	100(51.2%)

In 1 decade and 2 decade with respect to indication for OH, the contribution of rupture uterus reduced significantly (61% vs 22%, p value < 0.001), atonic PPH increased slightly (19% vs 39%), traumatic PPH increased slightly (7% vs 15%), and morbidly adherent placenta increased significantly (7% to 23%, p value 0.007) (Table 3). Three cases of abdominal pregnancy (with placental adherence to posterior uterine wall and delayed OH due to vesicovaginal fistula) were categorized separately.

Table 3.

Indication of OH

Indication	2001–2010	2011–2021	Total	P value
Morbidly adherent placenta	5(7.14%)	29(23.2%)	34(27.2%)	0.007
Rupture uterus	43(61.42%)	28(22.4%)	71(36.41%)	< 0.001
Uterine atony	13(18.57%)	38(30.4%)	51(26.15%)	0.1
Traumatic PPH	5(7.14%)	19(15.2%)	24(12.30%)	0.12
Sepsis	4(5.71%)	7(5.6%)	11(5.6%)	0.95
Early pregnancy complication	5(7.1%)	11(8.8%)	16(8.2%)	0.8
Others	0	4(3.2%)	4(3.2%)	0.00
Total	70	125	195	–

95 patients (48.7%) of the 195 OH underwent a complete hysterectomy, while 100 patients (51.2%) underwent a partial hysterectomy. An increase in the incidence of total hysterectomy from 40% to 53.6% in the January 2011–December 2021 cohort of patients can be attributed to rise in the cases of adherent placenta spectrum disorder.

Almost all patients (100%) received a blood transfusion with nearly 50% of patients received a massive blood transfusion. 15% of patients experienced urinary tract injuries during surgery, mostly to the bladder with two cases of ureteric injury near the VUJ junction and one case of intestinal injury was also noted. 10% of patients with Acute Kidney Injury (AKI) needed postoperative hemodialysis care, while close to 30% of patients had DIC at the time of presentation and 70% of patients requiring ventilator support and critical care. 48.2% perinatal mortality rate was recorded overall (Table 4). There was no statistically significant difference between the first and 2nd decades of the study period in terms of complications like maternal deaths, Disseminated Intravascular Coagulopathy (DIC), AKI, and urinary tract injuries. However, the quality of critical care improved (54% vs. 79% (p < 0.001), along with decreased wound infection rates and significantly lower perinatal mortality (61.9% vs 48%, p < 0.03) (Table 4).

Table 4.

Complication of OH

Complication	2001–2010(n = 70)	2011–2021(n = 125)	Total	P value
Urinary bladder injury	12(17.14%)	19(15.2%)	30(15.38%)	0.7
Blood transfusion	70(100%)	125(100%)	195(100%)	1
Massive Blood transfusion	40(57.14%)	61(48.8%)	101(51.79%)	0.4
CCU admission	38(54.28%)	99(79.2%)	137(70.25%)	0.04
Ventilator support	16(22.85%)	88(70.4%)	104(53.33%)	< 0.0001
DIC	24(34.28%)	33(26.4%)	57(29.23%)	0.3
Renal failure and Dialysis	8(11.42%)	9(7.2%)	17(8.71%)	0.3
ReLaparotomy	7(10%)	10(8%)	17(8.71%)	0.6
Wound infection including Burst abdomen	10(12.85%)	5(4%)	14(7.1%)	0.03
Perinatal mortality	44(61.97%)	50(48%)	94(48.20%)	0.03
Maternal death	11(15.71%)	25(20%)	36(18.46%)	0.5

Despite receiving life-saving OH, maternal mortality in severe PPH was seen due to DIC and an irreversible hypovolemic shock. In our study, 26 patients (72%) died due to underlying overt DIC and Nine deaths (25%), due to sepsis. The incidence of maternal death was similar in both time periods (16% vs 20%). DIC, hypovolemic shock, sepsis, subsequent multi-organ failure were the main documented causes of mortality (Table 4).

Discussion

The incidence of peri-partum hysterectomy is 2.8 per 1000 births in low and middle-income countries and 0.7 per 1000 deliveries in high-income countries, according to a systematic review and meta-analysis. One in a thousand births is complicate and leads to hysterectomy (0.2–10.1) [5]. The incidence of an obstetric hysterectomy varies among European nations from 0.5 to 1.076 [6]; in Dublin and Ireland, it is on the decline from 0.4 to 0.2 [7]. The frequency varies from 0.4 to 2.0 per 1000 live births in developing nations including Nepal, Africa, and India [8–10].

Since we are the only referral facility for the nearby rural population, a high influx of referred patients in their postpartum period presented to us, which were not included in the denominator (live births), leading to an extremely high incidence of OH about 9.9 per 1000 live births.

Cesarean Section and Obstetric Hysterectomy

The incidence of OH rises in tandem with the rate of cesarean sections. A statistically significant correlation between the growing CS rate and obstetric hysterectomy was established in a larger population-based European study [6] with similar trends observed in our study with a risen in the CS rate from 37 to 49% and OH from 8 to 14 per 10,000 live births in the 2 decade of the study timeline. Ageing, parity, a lack of prenatal care and unscheduled patients, as well as prior cesarean section experience, were identify as risk factors for hysterectomy [5]. In our analysis, we discovered that the majority of patients were in the 21–30 age range, had several pregnancies, and were referred from other centers that provided inferior prenatal care. There are many studies comparing the Double layer (DL) uterine closure technique versus with Single layer (SL) closure techniques during cesarean section. The risk of uterine dehiscence or rupture in patients with previous cesarean section is comparable in those undergoing single versus double layer closure [17, 18].

Parity and Obstetric Hysterectomy

The conservative treatment of PPH is now possible due to improvement of uterotonics, availability of blood components, and surgical competence as a result of advancement in modern obstetrics, which has led to a decline in OH rates among primigravida and the younger population [11]. In our study, the prevalence of OH dropped from 34 to 15% in primigravida, which was statistically significant (p < 0.05).

Indication of Obstetric Hysterectomy

As CS rates increases, morbidly adherent placentas are becoming a more frequent rationale for OH; at the same time, uterine rupture has dramatically decrease as a result of judicious oxytocic administration and a low propensity for the trial of labor after CS (TOLAC). Similar results were found in a more extensive population-based investigation conducted over a 40-year period by Flood KM et al [7]. The majority of recent studies on OH shown that, after uterine atony, the abnormally invasive placenta was becoming a more frequent cause of OH. The percentage of OH with abnormally adherent placentas varies from 21 to 60% [6–10, 14]. Similar results were also reported in our study. 7% to 23% of patients underwent OH were found to have abnormally adherent placentas (Table 5). Additionally, there had been a steady rise in the number of patients needing obstetric hysterectomy for traumatic PPH, most likely as a result of poor training in instrumental vaginal births throughout residency programs [15]. In the 2nd decade of our study, it increased from 7 to 15%.

Table 5.

Comparison of different studies for OH

Name of study	Detail of study	Country	Incidence of OH per 1000 birth	Total number of delivery	Total number of OH	Incidence of Total OH	Main indications
Kallianidis, AF et al. [6]	Descriptive, multinational, population-based 2004–2016	All European countries	0.5 to 1.07	2 498 013	1393	–	Uterine atony (35.3%) and abnormally invasive placenta (34.8%)
Flood et al [7]	Retrospective:1966–2005,	Dublin Ireland	0.4 (0.2)	872,379	358	74%	Hemorrhage (34%, 30%); Placenta accreta(13%, 46.5%)
Chawla J et al. [8]	Retrospective, observational, 2006–2014	India	0.8	67,572	56	15%	Uterine atony (25%) and morbidly adherent placenta(21.4%)
Rana s et al [9]	Retrospective 2011–2016	Nepal	2.0	16,445	33	36%	Uterine atony 33% Placental cause 24%
Rawashdeh et al. [10]	Retrospective 2004–2019	Jordan	1.38	44,905	62	71%	Abnormally Adherebt placenta 44% and uterine atony 11.86%
Pettersen S et al. [15]	Retrospective study 2008–2017	Norway	0.4	96,313	42	–	Uterine atony (45.2%) and adherent placenta (52.4%)
Heitkamp et aL [16]	Prospective 2014–2015	South Africa	1.4	32,862	59	–	Uterine atony 34% Invasive placenta 22%
Present study	Retrospective:2001–2021	India	9.9	19,666	195	53.6%	Uterine atony(19%,30%) Densly adherent placenta (7%,23%)

Placenta Accrete and Obstetric Hysterectomy

Placenta Accreta Spectrum (PAS) disorders can be diagnosed in the antenatal period by Ultrasonography. RI is used as a second—line screening tool for diagnosis of PAS. Loss of hypoechoic zone between the placenta and underlying myometrium, presence of multiple lacunae showing turbulent flow on Doppler, thinning of myometrium overlying the placenta and serosal invasion are some of the pathognomonic features of PAS on ultrasound [19, 20]. Rising incidence of PAS in repeat cesarean sections can be minimized by reducing primary CS rate as well as by educating the patient toward smaller family norms. Retrograde hysterectomy is a safer surgery in PAS patients as it involves posterior approach. The bladder dissection is done as a last step and by this time majority of the vessels have been ligated. The skill of approaching adherent placenta in a retrograde manner needs to be mastered [21]. Multidisciplinary approach involving interventional radiologist, vascular surgeon and senior obstetrician helps in reducing maternal morbidity and mortality in patients with PAS. Prophylactic internal iliac artery balloon occlusion prior to Cesarean section in PAS patients has shown promising results. It not only reduces life threatening hemorrhage but also reduces the chances of obstetric hysterectomy [22]^.

Total Vs. Subtotal Hysterectomy

The utility of total vs. subtotal obstetric hysterectomy is a topic of ongoing discussion. Recent trend of favoring complete hysterectomy may be related to an increase in adherent placenta occurrence. In a large study by Flood KM et al [7] & Rawashdeh et al. [10], nearly 70% of patients underwent a total hysterectomy compared of our study where 48% of population underwent the same surgery. In terms of complications, the requirement for ICU stays and predicted blood loss; complete and subtotal hysterectomy had similar outcomes [10].

Complication of Obstetric Hysterectomy

In our study's maternal mortality rate was 18.4%, which is similar to those by Chawla J et al. (17.9%) [8] and Rana et al. (12.1%) [9] but much higher than those by Flood et al. (1.1%) [7] and Rawshdeh et al. (1.6%) [10](Table 6). Hemorrhagic shock with underlying DIC was the leading cause of maternal death in both decades (72.2%). An interesting finding was a increased rate of coagulopathy of about 29.2% as compared to other studies [8, 10]. Delay in referral is linked to high mortality in patients with DIC and irreversible hemorrhagic shock; this also reflected as a major contributing factor to death (72% in both decades). High maternal mortality is connected with limited availability of blood and blood products, a lack of surgical expertise, low resources at the initial point of treatment in low- and middle-income countries [16].

Table 6.

Comparison of complication of OH

Name of study	Maternal deaths	Urinary track injury	DIC	Massive blood transfusion	Wound infection	Relaparotomy	ICU admission	Neonatal mortality	ARF
Chawla J et al. [8]	17.9%	–	12.5%	–	10.7%	3.6%	35.7%	28.6%	–
Rana s et al [9]	12.1%	3.0%	3.0%	–	–	6.1%	72.7%	45.5%	–
Rawashdeh et al. [10]	1.6%	27.1%	16.9%	–	3.38%	–	–	5.08%	3.3%
Pettersen S et al. [15]	–	9.5%	9.5%	61.9%	–	–	–	–	–
present study	18.46%	15.38%	29.2%	51.7%	7.1%	8.71%	70.25%	48.2%	8.7%

Urinary tract injury in our study was greater compared to Chawla et al [8] and Rana s et al [9], but lower compared to Rawashdeh et al [10]. Our study had 7.1%; wound infection rate was nearly identical to those of existing literature by Rana S. et al [8] and Chawla J et al [9]. Neonatal death and ICU admission rates in our study reached up to 70.2% and 48.2%, respectively, which was greater than Chawla J. et al [8] but similar to the study conducted by Rana S. et al [9]^. (Table 6).

AKI requiring dialysis was noted in about 8.7% patients in our study which was also higher then Rawashdeh et al [10]. This again can be explained due to late referral where irreversible hypovolemic shock damages renal function (Table 6). Maternal mortality, DIC, AKI, Urinary tract injuries remained static over 2 decades but Quality of critical care improved (54% vs 79% p < 0.001). Wound infection and perinatal mortality significantly decreased (p < 0.03). Maternal mortality remained static due to delay in seeking timely health care facilities and lack of a robust referral system.

Strength and Limitations of our Study

The Strengths of this study include: 1) A long period of time, roughly 20 years, were taken into account; 2) Trend analysis of OH performed, particularly in rural referral hospitals; 3) Other than indications of OH, many other parameters, such as critical unit stay, need for dialysis, and rate of Relaparotomy calculation, were studied as well; and 4) In-depth analysis of maternal death had been taken into consideration.

The Limitations of our study were: 1) It was a Retrospective study 2) There was a lack of precise blood transfusion data 3) Lack of data on the time interval between delivery to OH 4) No data regarding the number of centers visited prior OH 5) Majority of patients were referred as postpartum patients.

The high OH rate was owing to the fact that the majority of patients were referred in the postpartum period and were not included in the denominator (live births).

Conclusion

In order to prevent adherent placenta spectrum diseases, primary CS rates should be reduced. Maternal mortality can be reduced with effective prenatal care and the creation of a strong referral network. Appropriate use of skill laboratories and simulators will significantly improve the ability of future doctors to do surgical vaginal births. A multidisciplinary team approach, high-quality ANC treatment, and early detection of placenta accrete are instrumental to providing better care and to prevent avoidable surgical interventions.

Funding

This study not funded by any authority.

Declarations

Conflict of interest

All authors undersigned declare that they have no any conflict of interest.

Ethical approval

This article does not contain any animal studies. Ethical approval taken from HREC committee of institution before conducting this study. All authors declare that this article not under consideration for publication anywhere.

Informed consent

Written informed consent taken from all patients included in our study.