In the course of this work, we were confronted with the absence of certain information in the files.

During the study period, 14325 live births were registered at the Maradi CSME, and 379 cases of maternal deaths were noted. The maternal mortality ratio (MMR) was 2645.72 per 100,000 live births (LB).

Figure 1 shows that the highest rate was observed in 2019 with 3505.22 maternal deaths per 100,000 LB.

The mean age of the patients was 27.46 years, with extremes of 15 and 47 years. The 20-24 age group was the most represented with 88 cases (23.2%), followed by the 30-34 age group with 75 cases (20%). Housewives were 361 (95.3%), and 334 patients (88.1%) were uneducated. There were 261 of them from rural areas, i.e. 69%. The mean parity in our study was 3.46, with extremes ranging from 0 to 15. Grand multiparous cases (117 cases) (30.9%) accounted for the most significant proportion, followed by pauciparous cases (71 cases (18.7%). Only 42 patients (11.08%) had performed four or more Prenatal Consultations, and 171 patients (45.12%) had not performed any prenatal consultations ( Table 1 ).

Table 2 shows that the average distance covered per area is 84.6 kilometers(km) with extremes ranging from 4.4 to 164 km.

In this study, 363 patients (95.8%) were referred. The means of transport used was the ambulance in 85.8% of cases, i.e. 325 patients. 225 patients (59.4%) were admitted with an altered general condition, and 229 had pale conjunctivae (60.40%). Compared to the history, 369 patients (97.36%) had no medical history.

Table 3 shows that 264 deaths (69.65%) occurred in the post-partum period, 92 patients (24.27%) died ante-partum and 17 (4.48%) per-partum.

Of the 264 postpartum women who died in the postpartum period, 148 (56.06%) had given birth vaginally, 107 (40.53%) had undergone a caesarean section. The main indications for these cesarean sections were dominated by eclampsia (48.60%), retroplacental hematoma (24.30%), and dystocia (9.35%). The table above also shows us that 218 newborns (78.41%) were taken out alive, 49 newborns (17.63%) were stillborn, 11 newborns (3.96%) died after birth ( Table 4 ).

Figure 2 below shows that 215 cases (56.73%) of deaths were related to direct obstetric causes, and 164 cases (43.27%) were related to indirect causes.

As shown in Table 5 , direct causes of maternal mortalities were predominantly hemorrhage with 75 cases (19.79%), followed by hypertensive states with 73 cases (19.26%) and puerperal infections with 62 cases (16.36%). Indirect causes included anemia with 120 cases (31.66%), followed by malaria with 35 cases (9.24%).

Nearly half of the patients had died before 24 hours after admission, i.e. 49.86% ( Table 6 ).

Our study’s maternal mortality rate was 2645.72 per 100,000 live births (LB). Similar studies carried out in Niger a few years ago found rates of 3063.3 [3] and 3713.53 [4] per 100,000 live births, respectively. These very high intra-hospital rates reflect a national situation where national trends have remained high for several years. According to the WHO, it had risen from 491 deaths per 100,000 NV in 2015 to 441 deaths per 100,000 NV in 2020 [2] . Sujani Kempaiah and Coll. in India [5] in 2021 had found a mortality rate of 432.73 per 100,000 NV in their study, and Ketevi Ameya all in Togo [6] in 2023 had reported a rate of 555 deaths per 100,000 NV. We can explain our very high result because it is the only reference center that received all obstetric emergencies, often in deplorable conditions. This result further demonstrates the distortion between sub-Saharan and other countries in terms of maternal mortality. Elsewhere, Badis Bougali and all in Algeria [7] reported a maternal mortality rate of 101.3 deaths per 100,000 live births. According to the WHO, in 2020, sub-Saharan Africa was the only region with a very high maternal mortality ratio, estimated at 545 maternal deaths per 100,000 live births (uncertainty range 477 to 654) [2] .

In our series, the average age was 27.46 years, with extremes ranging from 15 to 47 years. It was 28.22 years, with extremes ranging from 15 to 44 years in the series reported by Fomulu et al. [8] in Cameroon. The average age in our study is lower than those of Philémon in the DRC [9] and Traoré in Ségou [10] , which found respectively 33 and 35.5 years old. This result is comparable to those of Garba M and Traoré, which reported 25% for the 20-24 age group and 26.1% for the 20-25 age group, respectively. In the series by Hansda J et al. in India [11] , in 2021, the most affected age group was 20 - 29 years old, with 59.7%. In our series, 95.3% of the patients who died were housewives, and 88.1% were uneducated. Housewives and uneducated women often lack income and depend financially on their husbands. This situation makes them vulnerable to decision-making about their health because they cannot afford the health costs. This socioeconomic aspect was reported in the Fomulu series, where 74.8% of the deceased patients were unemployed, and 64.08% had a level of education less than or equal to primary school [8] . In the Hansda J series [11] , 83.58% of the patients had a low socioeconomic status. The low socioeconomic level may be a contributing factor to maternal deaths, as it is women who have the most difficulty in consulting in time, thus aggravating their health status before any contact with the care teams. To this can be added the distance traveled by these women because 69% of these victims came from rural areas in our series. The most significant proportion of victims were large multiparous (30.9%), followed by pauci-parous (18.7%) and multiparous (18.2). It is a factor often linked to lifestyle, habits, and tradition. Several African studies indicate that the incidence of maternal mortality increases as we move from pauciparity to multiparity and from multiparity to great multiparity [9] . The greater incidence of large multiparous women in our series can also be explained by the pro-natalist context in the Maradi region, whose fertility rate per woman is the highest in Niger, at 7.1 during the study period. Also, large multiparous women are more predisposed to parturition and postpartum hemorrhages [12] [13] , which are among the leading direct causes of maternal death [9] —some authors such as Grover et al. In the PUNJAB in India [14] , in 2023, there was a predominance of maternal deaths of multiparous women in their series compared to parities.

In our studies, 69.65% of patients died in the postpartum period. Our result is higher than those of Garba M. et al. [4] and Philemon et al. [9] who found 60.1% and 58.88% of patients died in the postpartum period, respectively. Of these patients who died postpartum, 148 had given birth vaginally (56.06%) and 107 (40.53%) had given birth by caesarean section. The main causes of caesarean sections were eclampsia (48.60%), followed by retroplacental hematoma (24.30%). Laparotomy was performed in 9 patients (3.41%). Authors had reported higher rates of caesarean section among deceased patients ranging from 63.8% [15] to 66% [7] . But the link between caesarean delivery and the increased risk of maternal mortality compared to vaginal delivery had not been established. This high rate of caesarean sections is related to the severity of the maternal condition following the pathology that caused the death. In our series, a high percentage of women (24.27%) died during the antepartum period. This situation could be attributed to various factors, including delayed care-seeking, living in remote areas, and lack of awareness.

In our study, direct causes accounted for 56.73% of deaths, which is lower than the rates reported by Garba and Coll [4] and Badis Foughali [7] , who found 62% and 75.4% respectively. The WHO’s database for the period 2000 to 2020 shows that 75% of maternal deaths were due to direct causes. In developed countries with improved management of obstetric complications, maternal mortality has shown an increase in deaths from indirect causes compared to direct causes [7] . For example, in the UK, deaths from indirect causes (56.4%) exceeded those from direct causes (47.3%) [16] . The predominance of direct causes of maternal death in our study reflects sub-optimal management of obstetric complications in developing countries. Our analysis of the causes of death in our cohort shows a predominance of bleeding (19.78%), followed by hypertensive states (19.26%). These results are consistent with other studies that also found bleeding to be the primary direct cause of maternal death [4] [7] [17] . In our study, indirect causes accounted for a high percentage (43.27%) of maternal deaths, with anemia being the dominant cause at 31.66%, followed by malaria at 9.24%. The prevalence of anemia in our study can be attributed to the low socio-economic status and inadequate diet of the population, leading to chronic anemia during pregnancy. This anemia often results from undiagnosed bleeding, included in the anemia count at the time of death. Additionally, Niger is a country where malaria is endemic, with multiple consequences during pregnancy, including anemia. Regarding the length of stay before death, our study found that 49.86% of maternal deaths occurred within 24 hours of admission. This high rate can be attributed to the patients’ advanced deterioration upon reaching the referral center. This situation results from the often-circuitous patient journey, referral conditions, and especially the delay in seeking healthcare. Several authors have also reported the short time frame between admission and death, with very high proportions of fatalities (sources: [4] [11] [14] ). In our series, 69% of our patients came from rural areas and the distances covered by all patients varied from 4.4 to 164 km with an average distance of 84.6 km. This long circuit travelled by our patients before coming to the reference center could be a contributing factor in maternal deaths and also in the high proportion of deaths within 24 hours of admission.