Prevalence, characteristics, and outcomes of postpartum hemorrhage in a tertiary hospital in Tigray, Northern Ethiopia (2017–2021): retrospective descriptive study

Abstract

Background

Postpartum hemorrhage (PPH) is the major cause of maternal mortality and morbidity worldwide. Its burden is specifically significant in resource-limited settings.

Objective

The study aimed to assess the prevalence, characteristics, and outcomes of PPH in a tertiary hospital in Tigray, northern Ethiopia, between January 1, 2017, and December 31, 2021.

Methods

This was a cross-sectional study. Data on prevalence, sociodemographic, obstetrics, clinical characteristics, management, and outcome of cases diagnosed with PPH were analyzed using descriptive statistics.

Results

During the 5-year study period, 23,090 deliveries were recorded at Ayder Comprehensive Specialized Hospital (ACSH). A total of 269 cases of PPH were treated, providing a prevalence of 1.17% with 95%CI 0.027–1.30 (11 per 1000 deliveries). Twelve incomplete charts were excluded. Primary PPH accounted for 220 (85.6%) of the known cases. Most PPH (53.7%) were diagnosed within 2 h after delivery. The estimated documented mean blood loss was 1,572.9 ml (SD = 914.3). The mean hemoglobin level before PPH and after PPH was 12.2 mg/dl and 7.8 mg/dl, respectively. Regarding management, non-pneumatic anti-shock garment (NASG) was used in 79(30.7%) cases, misoprostol in 144 (56%) cases, oxytocin in 125 (48.6%), and ergometrine in 15 (5.1%) cases. Laparotomy was performed in 59 (23%) cases, and hysterectomy was conducted in 40 (67.8%) of these cases. Organ dysfunction (renal failure) was recorded in 34 (13.2%) cases, 14 (5.4%) developed coagulation dysfunction, and 17 (6.6%) deaths. Seven (3.1%) cases were admitted to the intensive care unit (ICU).

Conclusions and recommendations

The results of this study emphasized that PPH still poses a significant challenge for obstetric care in Tigray, Ethiopia. Considering the high case fatality rate, high prevalence of primary PPH, and atony, the practice of active management of the third stage of labor must be assessed, and corrective measures must be taken to institute proper protocols and practice.

Background

According to a report by the United Nations inter agency estimates (2023), 260,000 mothers globally die annually during pregnancy and childbirth. Postpartum hemorrhage (PPH) is the predominant cause of mortality and takes the largest share of the morbidities. Its burden is particularly significant in low-resource settings. Postpartum hemorrhage (PPH) remains a leading direct cause of maternal death worldwide; WHO notes that PPH results in about 70,000 maternal deaths each year and is a major contributor to severe maternal morbidity, particularly in low-resource settings [1, 2]. Though PPH has multiple definitions, the commonly utilized one is blood loss of 500 ml or more after vaginal delivery or more than 100 ml after cesarean section. It is further classified into primary PPH, which occurs within the first 24 h, and delayed PPH, which occurs between the first day and 6 weeks postpartum [3, 4].

The prevalence and characteristics of PPH vary according to the health infrastructure and resources. The prevalence is highest in low- and middle-income countries (LMICs) compared to developed countries. In LMICs, lack of medical supply, limited access to blood transfusion services, barred access to health facilities, high burden of anemia during pregnancy, and poor quality of childbirth services are the factors that contribute to a high burden of PPH. The lack of skilled delivery and poor postpartum follow-up further increase the risk of PPH, leading to severe maternal morbidities [5].

Ethiopia is recognized as one of the countries that takes a major share of the high burden of maternal mortality and morbidity. According to the Ethiopian Demographic Health Survey (EDHS), 2016, the maternal mortality ratio was 412 per 100,000 live births. The EDHS 2024–25 Key Indicators Report estimates a maternal mortality ratio of 141 maternal deaths per 100,000 live births representing marked decline of the maternal mortality. Though the decline is encouraging, it has long way to go to achieve the sustainable development goal target of less than 70. Furthermore, PPH persists to be listed as the leading cause of maternal mortality [6–9].

Primary PPH results primarily from uterine atony, where the uterus cannot contract effectively following delivery. Retained placental tissue, genital tract laceration, and coagulation disorders are the remaining causes. These fall under the “4T” framework mentioned in the recent guidelines. Delayed PPH can be due to several factors, including infection, subinvolution of the placental site, retained products of conception, and, rarely, gestational trophoblastic neoplasia. Understanding the causes of primary and delayed PPH is essential to developing targeted interventions, especially in low-resource settings where resources are meager. Multiple risk factors are attributed to the development of PPH [3, 10–12].

Studies show that PPH risk is increased by multiple maternal, obstetric, and intrapartum conditions. A large systemic review and meta-analysis published in 2025 showcased previous PPH, anemia, multiple pregnancy, placenta previa or abruption, caesarean section led to PPH. Furthermore, most guidelines highlight common obstetric contributors such as high parity, overly distended uterus, prolonged labor, and instrumental delivery lead to uterine atony and/or trauma. Underlying chronic medical conditions such as anemia and hypertensive disorders of pregnancy are repeatedly mentioned as relevant contributors in recent studies, including low resource setting. Early recognition of these factors is crucial for maternal health policy and clinical practice. Early risk detection allows health facilities to plan pre-emptive actions and readiness and to start early treatment. This allows them to undertake prophylactic measures (IV access, uterotonics, blood, and early escalation) and initiate early treatment to mitigate severe morbidities and mortalities [11, 13, 14].

The clinical presentation of PPH widely varies; the common features include heavy vaginal bleeding, which can be accompanied by signs and symptoms of hypovolemic shock such as hypotension, tachycardia, cold extremities, pallor, and dizziness. If the hypovolemic shock is not recognized or addressed, it can result in organ dysfunction, including renal failure, heart failure, and respiratory failure, as the body fails to distribute the limited volume of blood to vital organs. Disseminated intravascular coagulation (DIC) follows when the bleeding is massive [15, 16].

The management of PPH depends on the underlying etiology and severity of the bleeding. The first management line involves uterine massage, uterotonic drugs, and intravenous fluid administration. If these first- line measures do not work, additional interventions such as uterine tamponade, embolization of the uterine artery, ligation of the uterine artery, and compression sutures are instituted. If these subsequent interventions fail, invasive procedures such as a hysterectomy are required [17, 18].

In low-resource settings such as Tigray in northern Ethiopia, there is limited data on sociodemographic factors, clinical characteristics, complications, and management of PPH. This study aimed to assess the prevalence, characteristics, and outcomes of PPH in a tertiary hospital in Tigray, northern Ethiopia. The findings of this study will contribute to health policy and practice in low-resource settings.

Objective

The study aimed to assess the prevalence, characteristics, and outcomes of PPH in a tertiary hospital in Tigray, northern Ethiopia, between January 1, 2017, and December 31, 2021.

Methods

Study design and setting

A retrospective cross-sectional study was conducted at Ayder Comprehensive Specialized Hospital (ACSH) in Tigray, Northern Ethiopia, between January 1, 2017, and December 31, 2021. The hospital is a referral center for patients, including obstetric patients from Tigray and neighboring districts of the Afar and Amhara regions. The hospital has an ICU, and dialysis services and utilizes the blood bank of the region located in the city. The Department of Obstetrics and Gynecology has 50 beds and runs a residency program. The average annual number of deliveries is 6000. The department comprises 15 obstetricians and gynecologists, 40 residents, and 80 midwives and nurses. Maternity and delivery services are offered at no cost. The hospital sustains shortages of medical supplies and patients are requested to buy the missing supplies from private facilities.

Study population

The study included all women who were treated for PPH in the hospital during the study period. Women with incomplete medical records were excluded. Medical records with missing delivery summary or operation notes were excluded.

Because the study was retrospective descriptive, the sample size was determined by the number of deliveries and PPH cases recorded during the study period. However, we have calculated the power of the study using a precision-based approach for a single proportion. n = Z^2 p(1-p)/d^2 at 95% confidence (Z = 1.96). Using the expected PPH prevalence of 1.2% and an absolute precision of 0.5%, the required number of denominator is 1776 deliveries. The study included 23,090 deliveries, providing higher precision (95% margin of error ± 0.14% around the prevalence estimate).

Data collection

Data were extracted from medical records using a structured data collection form that was adopted from the literature published in a similar resource settings [10, 19, 20]. The study was conducted as part of a larger departmental initiative aimed at strengthening the research capacity of the Department of Obstetrics and Gynecology, including training faculty and clinicians in standardized clinical data abstraction. Data were collected by 11 obstetricians and gynecologists who received two days of training. Obstetricians/gynecologists were selected because key variables (PPH diagnosis and etiology, operative procedures, transfusion and critical care interventions, and complications) require specialist clinical interpretation from narrative documentation and operative notes, thereby improving the accuracy and completeness of extracted.

Study variables

The prevalence, etiologies, and classification of PPH were the primary outcome of interest. The risk factors evaluated included age, parity, mode of delivery, duration of labor, labor induction, place of delivery, delay, mode of delivery, phases, and pre-existing medical conditions. Secondary outcomes included the management of PPH, surgical intervention, blood transfusion, maternal morbidity and mortality.

Definition of terms

PPH was diagnosed according to the notes from the treating physician, which included a diagnosis made after genital tract bleeding of 500 ml after vaginal delivery or 1000 ml after hysterectomy. Primary PPH was defined as bleeding that occurred within the first 24 h after delivery. Delayed PPH was defined as bleeding that occurred after 24 h but before 12 weeks postpartum [11, 21].

The three delays framework were defined as follows: delay I (delay at home/ decision to seek care) was recorded when the medical record documented late decision making or prolonged stay at home after onset of vaginal bleeding. Delay II (delay in reaching health facility/ delay in transport) was recorded when the medical record documented transportation related delay (prolonged travel time, lack of transportation, delay in referral transfer). Delay III (delay at the health facility) was recorded when the medical records documented delay after arrival at the health facility before assessment or initiation of PPH management (delay in evaluation, delayed availability of drugs or blood products, delayed uterotonic or surgical intervention). For all phases, delay was documented from clinical notes, referral notes, and coded as Yes/No.

Data analysis

Data were entered and analyzed using SPSS version 25 (IBM Corp., Armonk, NY, USA). Descriptive statistics were calculated for all variables. Frequencies, percentages, and 95% confidence intervals (CIs) were calculated for categorical variables. Continuous variables were summarized using mean (SD) or median (IQR), as appropriate. The prevalence of PPH was calculated by dividing the number of PPH cases by the total number of deliveries during the study period. No bivariate or multivariable analyses were conducted; therefore, no hypothesis-testing statistical tests or model-based significance levels were applied. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)Statement: guidelines for reporting observational studies was utilized to prepare this manuscript [22].

Ethical considerations

Ethical approval was obtained from the Institutional Review Board of Mekelle College, College of Health Sciences. This is part of a large study on the magnitude, trend, and causes of maternal near misses and mortality at ACSH (approval number: MU-IRB 1950/2022). The ethical clearance covered abstraction of records from the study period January 1, 2017, to December 31, 2021. Patient confidentiality was preserved by anonymizing the data during analysis, data were securely stored and used for research purposes only.

Results

During the study period, there were 23,090 deliveries at ACSH. A total of 269 cases of PPH were treated, providing a prevalence of 1.17% with 95%CI 0.027–1.30 (11 per 1000 deliveries). Twelve incomplete charts (4.46%) were excluded. Thus, the analysis included 257 records. (Fig. 1).

Fig. 1.

Flow chart of total deliveries and PPH cases in ACSH from January 1, 2017, and December 31, 2021

Socio demographic characteristics and obstetrics characteristics

The mean age of the mothers was 29 years (SD = 6.3) (range 15–49 years). The majority, 82 (31.9%), fell within the age group of 25 to 29 years. This was followed by 65 (25.3%) aged between 30 and 34 years and 50 (19.5%) in the age group of 20 to 24 years. Among the mothers, 54.9% lived in rural areas. (Table 1).

Table 1.

Sociodemographic characteristics of mothers with PPH (N=257), 2017 – 2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

	Frequency(n=257)		95% CI of percent
Characteristics		Percent	Lower limit	Upper limit
Residence
Urban	116	45.1%	39.1%	51.2%
Rural	141	54.9%	48.8%	60.9%
Age in years (Mean=29.0, SD=6.3)
<20	9	3.5%	1.8%	6.6%
20-24	50	19.5%	15.1%	24.8%
25-29	82	31.9%	26.5%	37.9%
30-34	65	25.3%	20.3%	31.0%
35-39	28	10.9%	7.6%	15.4%
40-44	18	7.0%	4.5%	10.9%
45-49	5	2.0%	0.8%	4.6%

ANC Antenatal care

Obstetric characteristics

The mean gravidity was 3.67 + 2.53 (1-11), while the mean parity was 3.41 + 2.48 (1-11). Almost half of the mothers, 119 (46.3%), were multiparous (para 2–4). Followed by grand multiparous (para > 5) 69(26.9%), and primiparous (para 1) 69 (26.9%).

Nearly half of the index pregnancies, 124 (48.2%), were dated. Most mothers were at term (37–41 weeks), 64 (51.6%), followed by post term (> 42 weeks), 31 (25%), and late preterm (34–36 weeks) 18 (14.5%). Most of the mothers, 177 (68.9%), attended ANC, of which most of them 100 (56.5%) visited health centers. (Table 2).

Table 2.

Obstetric characteristics of mothers with PPH (N=257), 2017 – 2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

	Frequency(n=257)		95% CI of percent
Characteristics		Percent	Lower limit	Upper limit
Gravidity
Primigravida	55	21.4%	16.8%	26.9%
Multigravida	126	49.0%	42.9%	55.2%
Grand multigravida	76	29.6%	24.3%	35.5%
Parity
Nulliparous	1	0.4%	0.1%	2.7%
Primiparous	68	26.5%	21.4%	32.2%
Multiparous	119	46.3%	40.3%	52.5%
Grand multiparous	69	26.9%	21.8%	32.6%
Number of abortions
None	229	89.1%	84.9%	92.5%
Once	22	8.6%	5.6%	12.4%
Twice	6	2.3%	1.0%	4.7%
Is gestational age known?
Yes	124	48.2%	42.2%	54.3%
No	133	51.8%	45.7%	57.8%
Gestational age category (n=124)
<34 weeks	11	8.9%	5.0%	15.4%
34-36 weeks	18	14.5%	9.3%	22.0%
37-41 weeks	64	51.6%	42.8%	60.4%
42+ weeks	31	25.0%	18.1%	33.5%
Did the mother attend ANC
Yes	177	68.9%	63.0%	74.3%
No	80	31.1%	25.7%	37.0%
Place of ANC visit (n=177)
Health center	100	56.5%	49.1%	63.6%
Primary Hospital	25	14.1%	9.6%	19.8%
General hospital	16	9.0%	5.5%	13.9%
Specialized Hospital	30	16.9%	12.0%	23.0%
Unknown	6	3.4%	1.4%	6.9%

Obstetrics and medical complications of the index pregnancy

Of the obstetric complications of pregnancy in the index pregnancy, 31 (12.1%) of the mothers had previous cesarean section (CS), 31 (12.1%) had a uterine rupture, preeclampsia 22 (8.6%), placental abruption 21 (8.3%), multiple pregnancies 18 (7%), and placenta previa 15 (5.8%). In terms of medical complications during the index pregnancy, 73 mothers (28.4%) had anemia, 8 (3.1%) had chronic hypertension, and 5(1.9%) had overt diabetes. (Table 3).

Table 3.

Obstetrics and Preexisting Medical Complications of the index pregnancy of mothers with PPH (N=257), 2017 – 2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

	Frequency(n=257)		95% CI of percent
Characteristics		Percent	Lower limit	Upper limit
Abruptio placenta
Yes	21	8.2%	5.3%	12.0%
No	236	91.8%	88.0%	94.7%
Placenta previa
Yes	15	5.8%	3.5%	9.2%
No	242	94.2%	90.8%	96.5%
Chronic hypertension
Yes	8	3.1%	1.5%	5.8%
No	249	96.9%	94.2%	98.5%
Preeclampsia/Eclampsia
Yes	22	8.6%	5.6%	12.4%
No	235	91.4%	87.6%	94.4%
Anemia
Yes	73	28.4%	23.2%	34.1%
No	184	71.6%	65.9%	76.8%
Polyhydramnios
Yes	4	1.6%	0.5%	3.7%
No	253	98.4%	96.3%	99.5%
HIV status
Yes	3	1.2%	0.3%	3.1%
No	185	72.0%	66.3%	77.2%
Unknown	69	26.8%	21.7%	32.5%
Overt diabetes mellitus
Yes	5	1.9%	0.7%	4.2%
No	252	98.1%	95.8%	99.3%
Gestational diabetes mellitus
Yes	5	1.9%	0.7%	4.2%
No	193	75.1%	69.5%	80.1%
Unknown	59	23.0%	18.1%	28.4%
Previous cesarean delivery
Yes	31	12.1%	8.5%	16.5%
No	226	87.9%	83.5%	91.5%
Total	257	100.0%	.	.
Multiple pregnancy
Yes	18	7.0%	4.4%	10.6%
No	239	93.0%	89.4%	95.6%
Cardiac disease
Yes	2	0.8%	0.2%	2.5%
No	255	99.2%	97.5%	99.8%
Uterine rupture
Yes	31	12.1%	8.5%	16.5%
No	226	87.9%	83.5%	91.5%
Myoma
Yes	9	3.5%	1.7%	6.3%
No	248	96.5%	93.7%	98.3%

Characteristics of PPH

Half of the PPH 129 (50.2%) occurred at ACSH, while 102 (39.7%) occurred at other health facilities, and 26 (10.1%) occurred at home. Primary PPH accounted for 220 (85.6%) of the known cases. The majority of PPH 138 (53.7%) were diagnosed within 2 h of delivery.

Of the 257 cases of PPH, 195 (75.9%) were diagnosed during the weekdays, while 62 cases (24.1%) were diagnosed during the weekends. Taking into account the number of days, this provides an average of 39 cases per weekday (over five days) and approximately 31 cases per weekend day (over two days). Furthermore, in terms of time of day, 192 (74.7%) were during the day, while 92 (35.8%) cases were diagnosed during the night. The specific time of occurrence was not recorded for 36 (14%).

The duration of labor was known only for 104 (40.5%) cases. The median duration of labor was 8.5, with a quartile range of 11. Labor was induced or augmented for 23 (8.9%) cases.

Many mothers, 158 (61.5%), delivered vaginally, followed by CS 94 (36.6%), Vacuum 3 (1.2%), and Forceps 2 (0.8%).

Most, 206 (80.2%) of the mothers, were referred to ACSH from other facilities. Half of the mothers, 133(51.8%), delivered in ACSH, while 124 (48.2%) delivered in other facilities.

The main sources of referrals were health centers 84 (40.8%), primary hospitals 62 (30.1%), and general hospitals 44 (21.4%). (Table 4).

Table 4.

Circumstances of PPH Diagnosis (N=257), 2017 – 2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

	Frequency(n=257)		95% CI of percent
Characteristics		Percent	Lower limit	Upper limit
Diagnosis of PPH with regard to time
Within 2 hours of postpartum	138	53.7%	47.6%	59.7%
>2 hours -12 hours of postpartum	62	24.1%	19.2%	29.6%
>12 hours-24 hours of postpartum	20	7.8%	5.0%	11.5%
>24 hours of postpartum	28	10.9%	7.5%	15.1%
Unknown	9	3.5%	1.7%	6.3%
Where did the PPH occurred
Ayder	129	50.2%	44.1%	56.3%
Other health facility	102	39.7%	33.8%	45.8%
Home	26	10.1%	6.9%	14.2%
Was the mother referred?
Yes	206	80.2%	75.0%	84.7%
No	51	19.8%	15.3%	25.0%
If referred place of referral (n=206)
Health center	84	40.8%	34.2%	47.6%
Primary hospital	62	30.1%	24.1%	36.6%
Private hospital	7	3.4%	1.5%	6.6%
General hospital	44	21.4%	16.2%	27.3%
Specialized hospital	1	0.5%	0.1%	2.2%
Other	8	3.9%	1.9%	7.2%
Place of delivery
Ayder	133	51.8%	45.7%	57.8%
Outside Ayder	124	48.2%	42.2%	54.3%
Mode of delivery
SVD	158	61.5%	55.4%	67.3%
Vacuum	3	1.2%	0.3%	3.1%
Forceps	2	0.8%	0.2%	2.5%
CS	94	36.6%	30.9%	42.6%
Is duration of labor known?
Yes (Median=8.5, IQR=11)	104	40.5%	34.6%	46.5%
No	153	59.5%	53.5%	65.4%
Was labor induced/ augmented?
Yes	23	8.9%	5.9%	12.9%
No	234	91.1%	87.1%	94.1%
If induced /augmented what was the indication? (n=23)
Labor abnormality	12	52.2%	32.5%	71.3%
Pregnancy related complications	11	47.8%	28.7%	67.5%
Indication for cesarean delivery (n=94)
Non reassuring fetal heart rate	17	18.1%	11.3%	26.8%
CPD	17	18.1%	11.3%	26.8%
CS scar + X factors	8	8.5%	4.1%	15.4%
Breech	4	4.3%	1.5%	9.8%
Transverse lie in labor	2	2.1%	0.4%	6.6%
Other	46	48.9%	39.0%	58.9%

CPD Cephalo Pelvic Disproportion, CS Cesarean Section, SVD Spontaneous Vaginal Delivery

Regarding delays, delays at home contributed to 78 (30.4%), delays in transport to 105 (40.9%) and delays at health facility 115 (44.7%). (Fig. 2).

Fig. 2.

Delays of PPH diagnosis (N = 257), 2017–2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

Clinical presentation of PPH

The mean estimated blood loss was 1,572.9 ml ± 914.3 ml. The mean hemoglobin level before the diagnosis of PPH was 12.2 g/dl ± 2.7 g/dL, while the mean hemoglobin level after diagnosis was 7.8 g/dl ± 2.7 g/dL. The mean recorded systolic blood pressure was 91.3 mm Hg ± 29.6 mm Hg, and the mean diastolic blood pressure was 56.7 mm Hg ± 20 mm Hg. Furthermore, the mean pulse rate was 109 beats per minute ± 24.6 beats per minute. (Table 5).

Table 5.

Clinical presentation of mothers with PPH at diagnosis (N = 257), 2017–2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

Statistics	SBP	DBP	PR	Temperature	Oxygen saturation (%)	Estimated blood loss (ml)	Hgb before PPH diagnosis	Hgb after PPH diagnosis
N	227	227	227	143	136	96	132	201
Mean	91. 3	56.7	109.0	36.3	92.5	1572.9	12.2	7.8
Sd	29.6	20.0	24.6	3.7	12.8	914.3	2.7	2.6
Min	0	0	0	0	0	0	0	1
Max	220	124	180	39.8	100	5500	22.1	16
Range	220	124	180	39.8	100	5500	22.1	15
P25	80	50	100	36.3	93	1000	11.2	6
P50	90	60	112	36.8	95	1500	12.4	7.8
P75	110	67	121	37	97	1975	14	9.7
IQR	30	17	21	0.7	4	975	2.9	3.7
Skewness	-1.0	-0.7	-1.7	-8.0	-5.3	1.3	-1.0	0.2

SBP Systolic blood pressure, DBP Diastolic Blood Pressure, PR Pulse rate, Hgb Hemoglobin

Etiological causes of PPH

Atony was the leading cause of PPH identified in 117 (45.5%), followed by genital tract injury 57 (22.2%), retained product of conceptus (RPC) 45 (17.5%) and retained placenta 40 (15.6%). Of the genital tract injuries, uterine rupture was the common cause, accounting for 24 (42.1%), followed by perineal injury 16 (28.1%) and uterine extension 14 (24.6%). Eight (20%) of the retained placenta were morbidly adherent. (Table 6).

Table 6.

Etiological Causes of PPH (N = 257), 2017–2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

Characteristics	Frequency (n = 257)		95% CI of percent
		Percent	Lower limit	Upper limit
Atony
Yes	117	45.5%	39.5%	51.6%
No	140	54.5%	48.4%	60.5%
Retained placenta
Yes	40	15.6%	11.5%	20.4%
No	217	84.4%	79.6%	88.5%
RPC
Yes	45	17.5%	13.2%	22.5%
No	212	82.5%	77.5%	86.8%
Genital tract injury
Yes	57	22.2%	17.4%	27.6%
No	200	77.8%	72.4%	82.6%
Adherent placenta (n = 227)
Yes	8	3.5%	1.7%	6.5%
No	219	96.5%	93.5%	98.3%
Genital tract injury-perineal (n = 57)
Yes	16	28.1%	17.7%	40.6%
No	41	71.9%	59.4%	82.3%
Genital tract injury: Cervical (n = 57)
Yes	10	17.5%	9.4%	28.9%
No	47	82.5%	71.1%	90.6%
Genital tract injury: Rupture (n = 57)
Yes	24	42.1%	29.9%	55.0%
No	33	57.9%	45.0%	70.1%
Genital tract injury: Uterine extension (n = 57)
Yes	14	24.6%	14.8%	36.8%
No	43	75.4%	63.2%	85.2%

Management and outcome

The median number of blood transfusions was 3, with an interquartile range (IQR) of 2. A total of 221 (86%) patients received a transfusion. In terms of uterotonic drug utilization for PPH treatment, misoprostol was used in 144 (56%) cases, oxytocin in 125 (48.6%), and ergometrine in 15 (5.1%) cases. Laparotomy was performed in 59 (23%) cases, and hysterectomy was conducted in 40 (67.8%) of these cases. B-Lynch sutures and uterine artery ligation were performed in 6 (10.2%) cases, while relaparotomy was performed in 18 (30.5%) cases. Among the temporary measures, the nonpneumatic antishock garment (NASG) was used in 79 (30.7%) cases and bimanual/aortic compression was used in 5 (1.9%) cases. Manual exploration was carried out in 51 (19.8%) cases, and curettage/manual vacuum aspiration was carried out in 13 (5.1%) cases. (Table 7).

Table 7.

Medical and surgical management of mothers with PPH (N = 257), 2017–2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

Characteristics	Frequency (n = 257)		95% CI of percent
		Percent	Lower limit	Upper limit
Blood transfused
Yes (Number of blood bugs: Median = 3 IQR = 2)	221	86.0%	81.4%	89.8%
No	36	14.0%	10.2%	18.6%
Misoprostol used
Yes	144	56.0%	49.9%	62.0%
No	113	44.0%	38.0%	50.1%
Oxytocin used
Yes	125	48.6%	42.6%	54.7%
No	132	51.4%	45.3%	57.4%
Ergometrine used
Yes	15	5.8%	3.5%	9.2%
No	242	94.2%	90.8%	96.5%
Laparotomy done
Yes	59	23.0%	18.1%	28.4%
No	198	77.0%	71.6%	81.9%
B-lynch (n = 59)
Yes	6	10.2%	4.4%	19.8%
No	53	89.8%	80.2%	95.6%
Uterine artery ligation (n = 59)
Yes	6	10.2%	4.4%	19.8%
No	53	89.8%	80.2%	95.6%
Relaparotomy (n = 59)
Yes	18	30.5%	19.9%	43.0%
No	41	69.5%	57.0%	80.1%
Hysterectomy (n = 59)
Yes (subtotal hysterectomy = 14, total hysterectomy = 26)	40	67.8%	55.2%	78.6%
No	19	32.2%	21.4%	44.8%
NASG (n = 227)
Yes	79	34.8%	28.8%	41.2%
No	148	65.2%	58.8%	71.2%
Curettage /MVA (n = 227)
Yes	13	5.7%	3.2%	9.3%
No	214	94.3%	90.7%	96.8%
Manual removal of placenta (n = 227)
Yes	51	22.5%	17.4%	28.2%
No	176	77.5%	71.8%	82.6%
Genital tear repair (n = 227)
Yes	32	14.1%	10.0%	19.1%
No	195	85.9%	80.9%	90.0%
Bimanual/ Aortic compression (n = 227)
Yes	5	2.2%	0.8%	4.8%
No	222	97.8%	95.2%	99.2%

MVA Manual Vacuum Aspiration, NASG Non-Pneumatic Anti Shock Garment

Organ dysfunction (renal failure) was recorded in 34 (13.2%) cases, while 14 (5.4%) developed coagulation dysfunction. Seven (3.1%) cases were admitted to the ICU. There was no record of death. (Table 8).

Table 8.

Outcomes of mothers with PPH (N = 257), 2017–2021, Ayder Comprehensive Specialized Hospital, Tigray, Ethiopia

Characteristics	Frequency (N = 257)		95% CI Percent
		Percent	Lower limit	Upper limit
Organ dysfunction (Renal failure)
Yes	34	13.2%	9.5%	17.8%
No	223	86.8%	82.2%	90.5%
Coagulation dysfunction
Yes	14	5.4%	3.2%	8.7%
No	243	94.6%	91.3%	96.8%
ICU admission (n = 227)
Yes	7	3.1%	1.4%	6.0%
No	220	96.9%	94.0%	98.6%
Death
Yes  No	17 240	6.6% 93.4%	3.6% 90.4%	9.6% 96.4%

ICU Intensive Care Unit

Discussion

The study aimed to assess the prevalence, characteristics, and outcomes of PPH in a tertiary hospital in Tigray, northern Ethiopia, over a period of five years. The study showed a prevalence of PPH of 1.17% with 95%CI 0.027–1.30 (11 per 1000 deliveries). Primary PPH accounted for 85.6% of the known cases. Most PPH (53.7%) were diagnosed within 2 h after delivery. Half of PPH (50.2%) occurred at ACSH, 39.7% at other health facilities, and 10.1% at home. Atony was the leading cause of PPH identified in 45.5%, followed by genital tract injury (22.2%), RPC (17.5%), and retained placenta (15.6%). Delays at home contributed to 30.4%, transport delays to 40.9%, and delays at the health facility to 44.7%. Maternal death occurred in 6.6% of the PPH cases. Organ function dysfunction (renal failure) was recorded at 13.2%.

. The prevalence of PPH in this study, 1.17%, is within the range reported in more recent hospital-based African data. For example, a multicenter Ethiopian obstetric surveillance study (EthOSS, 2021–2022) reported an incidence of 8 per 1000 births (0.8%) in public hospitals in eastern Ethiopia, which is close to our estimate [10]. At the continental level, a recent meta-analysis of African studies (2003–2023) reported a pooled PPH prevalence of 4.2%, with Eastern Africa around 1%, aligning with our finding [23]. In contrast, a large nationwide Nigerian referral-hospital network (2019–2020) reported a higher prevalence of 3.2% [24]. Furthermore, the findings were much lower than the prevalence reported by a systematic review conducted in Africa, which ranged from 7.2% to 25% depending on the region, and a systematic review conducted in Ethiopia, which reported the prevalence of 8.2%. Another study conducted in Cameroon by Hale-Ekane et al. reported a prevalence of 23.6%, which is the highest reported in Africa [7, 25–27]. This wide variation showcases regional differences in healthcare systems, access to maternal care, differences in the patient population, the magnitude of risk factors, and methodologies. The low rate observed in this study could be attributed to high ANC coverage and high institutional delivery in the Tigray Region [28].

The findings of this study indicate that primary PPH accounted for 85.6% of the known cases. Of which most of them were diagnosed within 2 h after delivery. This highlights the importance of early diagnosis and intervention. This was higher than the findings of Halle-Ekane et al., who reported that half of the PPH cases occurred within the first day of delivery. Early diagnosis and management significantly reduce maternal morbidity and mortality [29].

The predominant etiologic cause of PPH in this study was uterine atony, accounting for 45.5% of cases. This is consistent with global tendencies to mention atony as the most common cause of PPH. In a large systematic review and meta-analysis uterine atony accounted for the largest share of reported PPH causes. In low-resource settings, uterine atony is frequently seen in association with factors such as high parity and prolonged or obstructed labour, which are also recognised contributors in recent summaries of PPH risk [14]. Astudy from Ethiopia reported that 77% of PPH cases were due to uterine atony. These studies have shown the importance of addressing and preventing uterine atony. Active management of the third stage of labor, which comprises the administration of a uterotonic agent after delivery, controlled cord traction, and uterine massage, has been shown to reduce the risk of atony and subsequently reduce PPH by 60% [10, 16, 30]. The status of whether the cases in this study received AMTSL is not recorded, so it remains unknown.

Comparatively, genital tract injuries were responsible for 22.2% of PPH cases in this study. This is consistent with pooled evidence showing that genital tract trauma is an important contributor to PPH, after uterine atony, alongside retained placenta and coagulation disorders [23]. Recent multicenter surveillance data from Ethiopia (EthOSS, 2021–2022) similarly report uterine atony as the main underlying cause of PPH, with other causes including genital tract trauma [10]. The difference between these settings shows the need for setting specific interventions and the need for comprehensive training for healthcare professionals in treating different causes of PPH.

RPC and retained placenta were reported as causes in 17.5% and 15.6% of the cases, respectively. The prevalence of the development of retained products of conception was reported to be between 1 and 6% of all pregnancies [31, 32]. RPC likely leads to severe PPH, if it is coupled with the placenta previa or with a morbidly adherent placenta. RPC and retained placenta are challenging in resource-limited settings due to meager resources and a lack of skilled health care providers. Furthermore, RPC contributes significantly to secondary PPH. A systematic review in Ethiopia indicated that RPC is the leading cause of delayed PPH [7, 33]. These findings underscore the importance of high-quality postpartum care and intervention.

The findings of the present study show a reduction in hemoglobin level from 12.2 g/dl to 7.8 g/dl after PPH, which is consistent with documented blood loss (mean estimated blood loss of 1,572.9 ml ± 914.3 ml). The documented mean systolic blood pressure (91.3 mm Hg ± 29.6 mm Hg) and diastolic blood pressure (56.7 mm Hg ± 20 mm Hg) were lower than the normal values, suggesting hypotension due to blood loss [15]. Furthermore, the mean pulse rate was 109 beats per minute, ± 24.6 beats per minute, suggesting a compensatory reflex to bleeding [34]. Variation is observed when these findings are compared with studies done elsewhere. A study conducted in southern Ethiopia reported an average estimated blood loss of 1,000 ml with a hemoglobin drop of 8.5 g/dl. Another study from Africa demonstrated an average blood loss of 1,200 ml with a hemoglobin drop of 9 g/dl. The difference in values is attributed to differences in the health system, resources, and the magnitude of risk factors. The difference in blood volume and hemoglobin levels is influenced by the techniques used to measure blood loss. The American College of Obstetricians and Gynecologists recommend the use of Quantitative blood loss measurement, a more precise assessment compared to subjective report. However, the reported outcomes are hugely affected globally due to variation in the application of the methods [15, 35, 36]. In this study, a subjective evaluation of blood loss was recorded.

The results of this indicate the huge burden of PPH and the different approaches to its management. Blood transfusions are very important aspects of the treatment of PPH [37]. Most of the patients (86%) were transfused. Furthermore, the median number of transfusions was 3, indicating that multiple cases required multiple transfusions. Several reasons can explain this high rate of transfusion. First, the mean estimated blood loss was 1,572.9 ml. Studies have indicated that massive bleeding is associated with an increased chance of transfusion. Second, 28.4% of the cases in the present study had pre-existing anemia. Pre-existing medical conditions, such as anemia, have been proven to be a risk of a need for transfusion.

The three delays were common in this study; 30.4% of cases had delays at home, 40.9% had delays in reaching a health facility,45% experienced Delay III (after reaching the health facility). Using the three delays model, several plausible contributors may explain these findings in low resource settings. Delay I (decision to seek care) is usually influenced by limited knowledge of danger signs of pregnancy, low autonomy/decision-making power, and financial constraints. Community health education and birth-preparedness interventions have been shown to improve knowledge of danger signs and preparedness practices, and male partner involvement can support timely care-seeking where men influence household decisions and resources [38–40]. Delay II (reaching care) commonly reflects transport, distance, referral coordination, and communication barriers; evidence suggests emergency transport interventions may reduce referral delay and improve access to emergency obstetric care [41]. Delay III (receiving timely care) is typically driven by facility readiness. Rapid recognition, protocol adherence, availability of uterotonics and supplies, timely blood products, theatre capacity, and staff skills. These are the areas emphasized in WHO guidance for PPH management and emergency obstetric care capacity monitoring. Delay in the institution of management is associated with more bleeding and subsequent need for transfusions [13]. However, it must be noted and interpreted that delay variables in this study were obtained from retrospective documentation instead of recorded timestamps. Thus, these findings must be noted as documentation-based indicators rather than exact measures of time of care. Future prospective and qualitative studies are required to identify and target specific modifiable bottlenecks across the three phases.

Misoprostol, oxytocin, and ergometrine were used in many patients. Additionally, temporary methods such as NASG were implemented. This aligns with the WHO recommendations as a non-invasive first-line treatment approach for PPH [42]. However, the study reported higher rates of laparotomy (23%) and subsequent hysterectomies (67.8% of laparotomies). This indicates that surgical interventions remain an important component in the management of severe cases. Uterine artery ligation techniques have a success rate close to 93% and 97% reported by the studies by Doumouchtsis et al. and O’Leary respectively [43, 44]. The low rates of B-Lynch sutures (10.2%), uterine artery ligation (10.2%), and bimanual/aortic compression (1.9%) in the present study indicate a need for further training and resources to implement these methods. Furthermore, the use of manual exploration (19.8%) and curettage/manual vacuum aspiration (5.1%) emphasizes the ongoing need for skilled manual interventions in PPH management.

Regarding the outcomes of PPH, this study reported maternal death of 6.6%, an organ dysfunction (renal failure) of 13.2%, while 5.4% developed coagulation dysfunction, and 3.1% of cases were admitted to the ICU. The PPH case fatality rate in our study (6.6%) is high when interpreted against accepted emergency obstetric care benchmarks. In EmONC monitoring, the maximum acceptable direct obstetric case fatality rate in facilities is < 1%, and values above this level suggest important gaps in timely and effective management and/or facility readiness. A similar pattern has been reported in recent Ethiopian multicenter study, where PPH accounted for 27% of hospital maternal deaths and the PPH case fatality rate was 6 per 100, with the authors highlighting suboptimal coverage of uterotonics and limited blood transfusion as plausible contributors [10, 45]. Regarding the other complications, these results mirror the reports of studies conducted in low-resource settings, where PPH is the leading cause of organ dysfunction and coagulopathies. A systematic review conducted by Ahmed Saad and Shemies reported PPH as the second common cause of renal failure in pregnancy following preeclampsia and eclampsia [46]. Similarly, a retrospective chart review in ACSH done by Berhe et al. reported that 40.6% of the cases were related to hemorrhage [47]. The magnitude of organ dysfunction reported in developed countries is generally lower than that reported in low-resource settings [48]. High prevalence of antenatal anemia, lack of adequate health care facilities, restrictions in accessing comprehensive obstetric care, and lack of skilled birth attendants lead to more morbidities.

This study has several limitations. First, because it was a retrospective chart review, the analysis depended on the completeness and quality of routine documentation. Some socio-demographic variables (such as education, marital status, and economic status) were not consistently recorded and could not be analyzed. Second, delays were identified from what was written in the medical records; we could not explore the underlying reasons for late diagnosis or delayed treatment, particularly for the 46.3% of cases diagnosed more than two hours after delivery. Addressing these causes will require future prospective and/or qualitative study. Third, the study is descriptive, so it cannot determine associations or causal relationships between potential risk factors and outcomes. Fourth, the findings reflect a single tertiary hospital and may not be generalizable to other settings. Finally, since the data cover 2017–2021, ongoing surveillance and periodic audits are needed to assess whether PPH patterns and health-system delays have changed in the current context. Despite these limitations, the study provides detailed clinical and management data on PPH in a tertiary referral setting and highlights key gaps relevant for quality improvement.

Conclusion

The results of this study emphasize that PPH posed a significant challenge for obstetric care in Tigray, Ethiopia. Primary PPH is the most common type of hemorrhage. The high prevalence of uterine atony as the leading cause of PPH, a large drop in hemoglobin following PPH, substantial mean blood loss, high maternal death rates, and high rates of organ dysfunction such as renal failure among the cases reviewed in this study all underscore the need for target-specific interventions. An effective intervention that was demonstrated to reduce PPH by multiple studies done elsewhere is the AMTSL [16, 30, 49]. Furthermore, tackling risk factors related to uterine atony, such as anemia, through high-quality antenatal care and skilled delivery can mitigate the comorbidities associated with PPH.

The study also found low rates of uterine-conserving procedures, such as B-Lynch sutures and uterine artery ligation, as well as bimanual or aortic compression, alongside high rates of laparotomy and subsequent hysterectomy. This indicates a need for improvement in healthcare infrastructure and training for healthcare providers in managing uterine atony, genital tract injuries, and retained products of conception (RPCs). Improving these aspects could lead to better maternal health outcomes in resource-limited settings.

Abbreviations

ACSH

Ayder Comprehensive Specialized Hospital

AMTSL

Active Management of the Third Stage of Labor

ANC

Antenatal Care

CI

Confidence Interval

CPD

Cephalo-Pelvic Disproportion

CS

Cesarean Section

DBP

Diastolic Blood Pressure

DIC

Disseminated Intravascular Coagulation

EDHS

Ethiopian Demographic Health Survey

HGB

Hemoglobin

ICU

Intensive Care Unit

LMICs

Low-and Middle-Income Countries

NASG

Non-pneumatic Anti-Shock Garment

PPH

Postpartum Hemorrhage

PR

Pulse Rate

RPC

Retained Product of Conceptus

SBP

Systolic Blood Pressure

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

SVD

Spontaneous Vaginal Delivery

Authors’ contributions

- **Conceptualization: ** Awol Yemane Legesse, Hale Teka- **Data Curation: ** Awol Yemane Legesse, Hale Teka, Mohamedawel Mohamedniguss Ebrahim, Marta Abrha, Daniel Gebre, Ephrem Berhe, Mohammedtahir Yahya, Fanus Gebru, Marta Abrha Gebru, Bisrat Tesfay, Hiluf Ebuy Abraha- **Formal Analysis: ** Mohamedawel Mohamedniguss Ebrahim, Hiluf Ebuy Abraha- **Investigation: ** Awol Yemane Legesse, Hale Teka- **Methodology: ** Awol Yemane Legesse, Mohamedawel Mohamedniguss Ebrahim, Hiluf Ebuy Abraha- **Project Administration: ** Awol Yemane Legesse, Hale Teka- **Writing – Original Draft: ** Awol Yemane Legesse, Hale Teka- **Writing – Review & Editing: ** All authors.

Funding

No funding was solicited for this study.

Data availability

All relevant data are within the paper.

Declarations

Ethics approval and consent to participate

Ethical approval for the study was obtained from the Mekelle University Institutional Review Board (approval number: MU-IRB 1950/2022) as part of a larger investigation into the magnitude, trends, and causes of maternal near misses and mortality at Ayder Comprehensive Specialized Hospital. As this study was a retrospective chart review, individual consent to participate was not required by the board.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.