Urinary infection (UI) corresponds to the attack on a tissue of the urinary tree by one or more microorganisms generating an inflammatory response, as well as signs and symptoms of variable nature and intensity depending on the background [1] . During pregnancy, it can manifest in several forms, the most serious of which is acute pyelonephritis of pregnancy (APN) [2] [3] . APN is a bacterial infection of the upper urinary tract (Pyelitis) and renal parenchyma (nephritis), complicating or associated with an infection of the lower urinary tract [4] .

The annual incidence of APN worldwide is 10.5 to 25.9 million [5] . The number of deaths caused by it is estimated at 4000 per year [6] . The risk of ANP is increased during pregnancy due to anatomical and physiological changes in the urinary tract [7] . The frequency of ANP during pregnancy ranges from 0.5% to 2% worldwide. In Algeria, 0.3% of pregnant women were hospitalized for APN [8] ; while Tchente et al. [9] reported a hospital prevalence of 0.6%. Several factors associated with the occurrence of APN have been found in the literature, notably sociodemographic and clinical factors. In California, black or Hispanic race, smoking, low education level and age less than 20 years were significantly associated with the occurrence of APN [10] . While in Jamaica, the age group between 20 and 29 was a risk factor [11] . Traore found that in Mali, the profession of housewife was significantly associated with the risk of occurrence of APN [12] . For some authors, the majority of APN occurred in the second trimester of pregnancy [11] [13] ; on the other hand, for others, they frequently occurred in the third trimester [14] . Additionally, some studies found that nulliparity was a risk factor while others showed an inverse relationship between parity and APN [13] [15] .

In Cameroon, few studies are available on this subject; for this reason, we proposed to conduct a case-control study in order to study the clinical and para-clinical profile linked to the occurrence of APN in the city of Yaoundé.

We used 236 files and retained 165 ( Figure 1 ).

We recruited in three teaching hospitals affiliated with the Faculty of Medicine of the University of Yaoundé 1 and as follows: 18 cases and 36 controls at the CHY; 22 cases and 44 controls at CHUY; 15 cases and 30 controls GOPHY ( Table 1 ).

The average age was 26.7 ± 5.3 (16 - 38) in cases and 26.5 ± 4.9 (16 - 36) in controls.

The mean gestational age at admission was 26.95 ± 7.91 (6.28 - 36.42) in cases and 27.48 ± 7.51 (7.42 - 37.14) in controls ( Table 2 ).

Cohabitation increased the risk of APN [OR: 2.19; 95% CI (1.03 - 4.62); P = 0.030] as opposed to age and level of education which were not associated with it ( Table 3 ).

* is for figures with statistical significance (the same as in the following tables).

Having sexual intercourse during pregnancy increased the risk of occurrence of APN [OR: 54.47; 95% CI (142.31 - 340.86); P = 0.01 as much as pre-gestational diabetes [OR: 10.9; 95% CI (1.44 - 261.43); P = 0.020] ( Table 4 and Table 5 ). However, there was no correlation between the trimester of pregnancy at admission and the occurrence of APN ( Table 6 ).

On the other hand, having had your first ANC after 15 weeks increased the risk of occurrence of APN by three times. [OR: 3.19; 95% CI (1.52 - 6.67); P = 0.001] while having done it before 15 weeks was a protective factor.

Having ANC in a district hospital increased the risk of occurrence of APN [OR: 7.4; 95% CI (2.3 - 27.49); P = 0.001]. Whereas, doing your ANC in a first category hospital was a protective factor.

Having your pregnancy monitored by a midwife/nurse increased the risk of occurrence of ANC [OR: 6.07; 95% CI (1.57 - 29.03); P = 0.01]. On the other hand, being followed by a gynecologist was a protective factor ( Table 7 ).

Pregnant women who had a genital infection during pregnancy had a 10 times higher risk of developing APN [OR: 9.94 95% CI (3.2 - 36.1); P = 0.001]. Having had a urinary infection during pregnancy increased the risk of developing APN [OR; 11.86 95% CI (3.87 - 42.6); P = 0.001]. Gestational diabetes increased the risk of occurrence of APN [OR; 5.37 95% CI (1.91 - 16.06); P = 0.001] ( Table 8 ).

In pregnant women with a urinary infection, the occurrence of fever, chills, lumbar pain, urinary burn and asthenia was significantly similar to APN ( Table 9 ).

In pregnant women with a urinary infection:

A normal temperature on admission was a protective factor for APN. A Heart Rate > 100 bpm [OR: 42.25; 95% CI (15.48 - 121.08); P = 0.001] and a FR > 20 cpm [OR: 4.85; 95% CI (2.34 - 10.25); P = 0.001] increased the risk of occurrence of APN. Fetal tachycardia increased the risk of finding APN [OR: 4.92; 95% CI (4.92 - 133.33); P = 0.001]. A normal fetal heart rate was a protective factor ( Table 10 ).

In pregnant women with a urinary infection:

Altered general state increased the risk of occurrence of APN [OR: 23.82; 95% CI (9.35 - 63.4); P = 0.001]. The Giordano sign was positive in 67.3% of cases on the right and in 32.7% of cases on the left. When they were in labor on admission, the risk of finding APN was increased [OR: 27.25; 95% CI (4.37 - 59.64); P = 0.001] ( Table 11 ).

We found no correlation between the organisms isolated at urine culture and the occurrence of APN. The germs most encountered in our population were:

Klebsiella pneumoniae: 12.7% in cases, and 10.9% in controls ( Table 12 ).

Sociodemographic characteristics

The latter being regularly reported in the literature as significantly associated with the occurrence of APN, they have been particularly studied.

Maternal age

The average maternal age was 26.7 ± 5.3 in the cases with extremes of 16 and 38 years. A look at studies carried out in the same geographical context would have given us more information on this. We unfortunately did not see any. However, our results are in agreement with the findings of Farkash et al. in Israel in 2012 where the average age of onset of APN was 26.3 ± 6.0 years [16] . Other authors such as Gang Jee Ko et al., on the other hand, have reported average ages higher or even lower than that of our study. In 2020, 31 ± 4 years for Gang Jee ko in a global cohort, while in Nepal and Texas, the average maternal ages were 22 ± 3.41 and 23.1 ± 5.1 years respectively [13] [17] . Unlike other authors such as Wing et al., we did not find a correlation between maternal age and the occurrence of APN. This data is found differently among all the authors we consulted. For Wing in California, women under 20 were more likely to have APN (OR = 2.0; CI = 95%, 1.8 - 2.3) [9] ; while Dawkins and Sharma the age range between 20 to 29 years is significantly associated with the risk of APN [11] [13] .

Level of education

In contradiction with literature data, we found no correlation between educational level and the occurrence of APN. However, some authors have dealt with this association in the literature, such as Wing et al., who found that compared to women who had 13 years of formal education, women who had only 12 years of education or less had a significantly increased risk of acute pyelonephritis (OR: 1.3; 95% CI: 1.2 - 1.5) [10] . Several other studies have found associations with low levels of education [18] - [20] . On the other hand, following Tchente et al., the high level of education was protective: primary level (P = 0.037; OR = 0.088; CI = 0.009 - 0.872) secondary level (P = 0.036; OR = 0.113; CI = 0.015 - 0.874) and university level (P = 0.03; OR = 0.106; CI = 0.014 - 0.81).

Marital status

This variable is variously reported by different authors regarding their association with APN. In our study we found that free union increased the risk of occurrence of APN [OR: 2.19; 95% CI (1.03 - 4.62); P = 0.030]. However, Labi et al. in their series found no link between marital status and APN [21] . Unlike Scholes in a case control study carried out in Washington which found an association between being single and the significant risk of having an APN [OR: 1.4; 95% CI (1.0 - 1.8); P = 0.005] [22] .

Clinical factors associated with APN

Parity

Like Tchente, we found no correlation between parity and the occurrence of APN [8] . However, from the work of Western authors, it appears that nulliparity increased the risk of occurrence of APN like Wing et al. who found that nulliparity was significantly associated with the risk of developing APN and represented 48.6% cases [10] . As did Hill and Farkash who reported that nulliparity represented 50% to 75% of cases of APN in their respective series [16] [18] .

For some, however, primiparity increased the risk of developing APN (OR [95% CI] 1.61 [1.55 - 1.67]) [17] . It should be noted that the average age of the pregnant women in these studies varied from 18 to 23 years; which is lower than the average age of our study which was 26 years and that of Tchente et al. which was 28 years.

Sexual intercourse

In our series, we found a correlation between sexual intercourse during pregnancy and the occurrence of APN (OR [95% CI] 54.47 [14.31 - 340.86]). These findings are in agreement with those of several authors in the literature including Fatton, Bethel, as well as Scholes et al. who described a link between sexual intercourse during pregnancy and APN [22] - [24] . Unlike Michel-Claire in a series in Clermont-Ferrand for whom no link existed between these two entities [25] .

The term of pregnancy

This data is reported differently by the literature consulted about. In our series, the mean gestational age at admission was 26.9 ± 7.9 weeks of amenorrhea (WA) in cases with extremes of 6 and 36 weeks, close to 27.7 ± 7.4 weeks in Saleh’s series [26] . But higher than the 22 ± 7.8 and 24.1 ± 4.7 WA found in the Dawkins and Zanatta studies; and lower than the 37.8 ± 6 found in Israel [11] [16] [27] . In our series, APN occurred in 90.9% of cases in the last two trimesters; this result is in line with the data in the literature because the authors agree on the fact that APN occurs most of the time during the last two trimesters of pregnancy; these two quarters also represent 90.8%, 90%, 88.5% of cases for some [10] [13] [28] ; and 79% of cases for others [18] . These results can be explained by the presence of certain hormonal and mechanical factors which predispose the pregnant woman to APN during the second half of pregnancy.

Prenatal follow-up

In our series, late initiation of ANC was weakly associated with the occurrence of APN while starting ANC before 15 weeks was a protective factor. The same goes for the Californian authors who, in an 18-year retrospective series, found that late initiation of prenatal visits increased the risk of having APN by 1.1 times (OR [95% CI] 1.4 [1.3 - 1.5]) [10] .

Diabetes mellitus

Diabetes is the bedrock of infections; this assertion is accepted by all [29] [30] . But in our series, we did not find an association between this metabolic disorder and the occurrence of APN; as much as Farkash reported in his series in Iran (P = 0.45) [15] . Pre-gestational diabetes, however, increased the risk of occurrence of APN (OR [95% CI] 10.9 [1.44 - 261.43]). Wing et al. confirmed this in their 18-year retrospective series where pre-gestational diabetes increased the risk of occurrence of APN by 1.7 times (95% CI, 1.3 - 2.1) [10] . Ditto for Hill et al. who, in Texas, pre-gestational diabetes mellitus increased the risk of having APN in the first trimester of pregnancy (P = 0.013) [15] While other authors like Dawkins, found no association between diabetes mellitus and the occurrence of serious urinary infections in pregnancy [11] .

History of urinary infection

In our study, a history of urinary infection during pregnancy emerged as a risk factor for APN (OR [95% CI] 11.86 [3.87 - 42.6]). Similar results are reported by Farkash (P = 0.001; OR = 10.3; CI = 4.8 - 22.1) [16] ; as well as by Gang jee ko in South Korea [17] .

HIV infection

HIV infection in our series with its deleterious effects on immune defense predisposed to a higher risk of developing APN (OR [95% CI] 5.52 [2.09 - 15.28]). This result is in agreement with data from the relevant international literature [31] [32] . The associated reduction in the immune system in addition, 20% to 30% of affected women may develop acute pyelonephritis [16] .

At the end of this research work, the objective of which was to study the factors associated with the occurrence of APM during pregnancy in three hospitals in the city of Yaoundé, it emerged that: free union, sexual intercourse during pregnancy, positive HIV serology, pre-gestational diabetes and a history of urinary infection during pregnancy were significantly associated with the occurrence of APN in pregnancy in Yaoundé.

The retrospective nature of our study imposes a certain number of inadequacies compared to that of the different authors, in particular the quality of the urine sample which was taken from pregnant women. Compared to those of other authors, the small size of our series, due to the deterioration of the files, could constitute a bias on our results due to a lack of statistical power.

Our study sheds light on this unexplored theme in our environment while opening the prospect of subsequent studies thus making it possible to develop operational strategies for prevention and early management of this condition with formidable impact for the mother and child.

Our gratitude to the hospital authorities as well as the various staff of these structures both for the authorizations and for the multifaceted supervision without which this retrospective study would not have been effective.

Essome and Mve supervised the study and writing of the manuscript, Mbog collected the data and wrote the manuscript, Tocki ensured the translation of the manuscript into English, as well as its formatting, Moustapha, Boten, Ofakem Ilick, Ndolo, Mangala, Ekono, Ngalame, Tchounzou, Ngaha, Ngono, Mounchikpou, Nana read and corrected the manuscript.

All authors have read and approved the final version.